Skip to content
Every release since May 13, 2026Major releaseUpdateSmall update

Gydr 4

8 releasesOctober 25 – November 18, 2026

  1. v4.3.2

    Latest

    This update fixes a few things you may have run into: audit logs that wouldn't open, a usage filter that wouldn't load, and a rare slow reply that left your visitor waiting too long.

    4 fixed

    Fixed4

    • Audit Logs open again — If your audit log included automatic actions, such as a billing update or a change made by the system, the Audit Logs page and its export could fail to load. Those entries now show alongside everyone else's.
    • Filtering usage by one credit type — Choosing a single credit type in your usage transactions could fail to load the list. It now filters as expected.
    • No long wait on a very slow reply — In rare cases a very slow reply left your visitor waiting over a minute before the chat said the message wasn't delivered. The chat now says so as soon as the reply can't be finished, with a Try again option.
    • More accurate usage reports — In rare cases your monthly usage report could count a credit twice. Usage still on its way is now counted once.
  2. v4.3.1

    This update makes Conversation Topics dependable for your longest conversations. A conversation with many messages is now sorted into its topics like any other, instead of going without any analysis, and nightly topic discovery keeps working through your conversations.

    3 fixed

    Fixed3

    • Long conversations now get their topics — A conversation with many messages that moved between several subjects could not be classified: it was retried several times and left without any analysis. It is now split into its topics like any other conversation.
    • Conversations that still can't be classified — If the AI model returns an answer that can't be used, the conversation is now recorded straight away under Other, with no credits charged, and its visitor goal and outcome still show in your dashboard.
    • Topic discovery stopping early — If the AI model couldn't name one group of conversations, discovery stopped for the night and tried again the next day. It now sets that group aside and carries on with the rest.
  3. v4.3.0

    The past few weeks have brought a wave of attacks on the software the whole web is built from: poisoned open-source packages, and serious flaws in widely used web frameworks. Gydr was not breached. But you trust us with your visitors' conversations, your catalogue and your account, so we treated that wave as a deadline. This release closes every critical security advisory in the components we rely on, pauses a suspended account completely, and keeps every secret out of your AI agent's chat. It also rebuilds every email Gydr sends, so the right people hear about the right things, and nobody else.

    7 new5 changed3 fixed3 breaking

    Breaking Changes

    These may need a change on your side.

    • Alert emails go to your team members — Alerts such as low credits are now sent to the members of your account who can act on them, and each person chooses which emails they get in Settings → Notifications. Extra email addresses typed into the old alert settings no longer receive them. Add those people to your team if they still need them.
    • Widget Studio has been removed — The visual canvas editor is gone. Every setting it offered is still available in your widget's settings.
    • Note for operators — This release changes what a Console sign-in token can carry (permission vocabulary generation bump — hash 92da18188af4). Once deployed, this release should not be rolled back on its own — doing so can briefly sign Console users out until the rollback is complete. If a rollback is ever needed, roll back together with the deployment layer that mints Console sessions.

    What's New7

    • Your AI agent can set up visitor sign-in, and you approve every step

      Your AI agent (Claude, ChatGPT and others, through Console MCP) can now prepare a full visitor sign-in setup: the sign-in source, the widget it protects and the connection that forwards a signed-in visitor's token. Nothing changes until you open the approve link in your own browser, check every value and approve it.

    • Secrets never pass through the chat

      When a setup needs a password, an API key, a header value or a test user's token, you type it on the Gydr approve page, never in the conversation with your agent. A test token works for one hour and is then deleted.

    • Every email, rebuilt

      Every email Gydr sends now shares one clean design that reads well in dark mode, says who sent it, and has a subject that tells you what it is about.

    • Unsubscribe in one click

      Emails you can switch off now carry an unsubscribe link, and your mail app's own Unsubscribe button works too. Emails about payments, security and access always arrive.

    • Help getting started, at the right time

      New accounts get a setup guide once their payment details are verified, a few gentle reminders while setup is unfinished, an introduction to connecting an AI agent, and a note if their chat goes quiet for a month. Each is sent once, and you can switch each off.

    • Payment receipts

      Every plan payment and credit pack purchase now sends a receipt showing the card used and who bought it.

    • Know when access changes

      A person removed from an account is told, and the account's team is told when someone leaves.

    Changed5

    • Security updates across the platform — We resolved every critical advisory and almost every other published advisory in the third-party components Gydr is built on, and moved to the latest security releases of our runtime and build tools.
    • A suspended account is fully paused — The chat widget, messaging channels, the Integration and Conversation APIs and AI agents all refuse a suspended account. It spends no credits and sends no emails until it is reactivated.
    • Notifications reach only the people who can act on them — Billing, credits, channels and other shared-area events go to the team members who can change that area, not to everyone who can view it.
    • Downgrades take effect at your next renewal — Your monthly credits stay as they are until the end of the billing period, and the lower price starts at your next renewal. No credit or refund is issued for the rest of the period.
    • Partner credit reclaims return only unused credits — When a partner reclaims, deactivates or detaches a managed account, monthly credits the account already used this month come back counted as used. The Reclaim review shows exactly how many.

    Fixed3

    • AI agent connections — Your agent now sees an account's real plan and status, including a partner-managed account's, and knows where to go to reactivate one.
    • Emails in dark mode — Headers and footers stay readable in mail apps that switch to a dark theme.
    • Notifications on phones — The notifications page is easier to read and use on a small screen.
  4. v4.2.0

    Your chatbot's replies now look and work more like your store: products, orders, carts and branches appear as cards your visitors can open and compare, and on a WooCommerce store your chatbot can now add items to a shopper's cart. Visitors can sign in through Supabase, and your chatbot sticks more closely to what your records actually say, in every language. This release also tightens security across the chat widget and the WordPress plugin.

    6 new8 changed10 fixed1 breaking

    Breaking Changes

    These may need a change on your side.

    • Reconnect AI assistants connected before this release — Apps connected to your Console before today keep working until December 13, 2026. Settings → Connected Apps marks them "Reconnect"; reconnect each one once and nothing else changes.

    What's New6

    • Richer replies in your chat widget

      Products a reply recommends now appear by name, each with a detail sheet, and two products can be compared side by side. Orders and carts from your store show as cards with a View order link. Branch cards show whether a branch is open now, its hours and its services. Replies can also show photo galleries and how many sources they used. When a visitor needs to sign in, the chat shows a short notice, with a Sign in button when your page tells the widget how to open your sign-in.

    • Your shopper's cart, in the chat (WooCommerce)

      On a WooCommerce store, your chatbot can now show a shopper's cart and add, change or remove items in it while they chat on your website, signed in or not. It works only on your own website, only on that shopper's own cart, and never places an order: checkout always opens your store's own checkout page. It is off until you turn it on, and needs the latest Gydr WordPress plugin and WooCommerce 11.1 or later.

    • Sign visitors in with Supabase

      Supabase now sits alongside your other providers under Settings → Visitor sign-in → Your keys. Type your project URL and Gydr fills in the rest. A visitor on an anonymous session stays a guest until they really sign in.

    • Test your Shopify connection

      A Shopify plugin now has a Test connection button. It tells you what is wrong and how to fix it.

    • Report a problem from your AI assistant

      An assistant connected to your Console can now report a problem with Gydr itself, naming the calls that went wrong. It can withdraw the report later if the problem turns out to be on your side.

    • WhatsApp channels that need a payment method

      When WhatsApp stops sending replies because your WhatsApp business account needs a payment method, Gydr now marks the channel, notifies you, and checks again every hour. No credits are spent on replies that cannot be delivered.

    Changed8

    • The connection fee follows the accounts you allow — Connecting an AI assistant uses a one-time fee in credits that now depends on how many accounts you tick, so connecting one account costs less. The fee is shown before you allow, and adding accounts later in Settings → Connected Apps uses no extra credits.
    • Safer app connections — The connect page warns you when an app is not verified, and disconnecting an app ends its access at once.
    • WordPress plugin 1.7.0 — Plugin updates are now signed and checked before they install. When you connect a site, it shows which Gydr account you are approving before the connection completes.
    • Local times on messaging channels — When a visitor on WhatsApp, Messenger, Instagram or Telegram confirms where they are, your chatbot uses that place's time zone for opening hours and times.
    • Sharper branch answers — Ask about one area and the chatbot lists only the branches there. Hours and services come straight from your records.
    • Clearer form errors in the Console — Plugin form errors now show on the field that caused them, on every tab, and the first one is focused for you.
    • Easier to use chat widget — Bigger tap targets, more readable text and working phone links. A field holding two phone numbers links each one separately.
    • Updated Terms of Service (v2.2) — Strengthened intellectual-property terms, and a dedicated address for reporting security issues: security@gydr.ai.

    Fixed10

    • Answers that stay with your records — Your chatbot no longer claims something about a group of products or branches that your records don't say. It keeps your FAQ's own wording. When your records answer a question, it gives that answer instead of a "not in our records" decline.
    • Questions in Chinese and other scripts — Questions and records written in a non-Latin script are now read correctly, so a good answer is no longer declined.
    • Products shown once, in the right place — Each product appears once, its card sits where the reply names it, and no product cards appear above an answer that comes from your FAQ.
    • Cart questions — A visitor asking about their cart gets their cart, not their past orders. Order cards show a completed date only for completed orders.
    • A reply that fails is not charged — If your chatbot can't answer, it tries once more. If that also fails, the visitor is told and no credits are used.
    • Signed-in visitors — A visitor whose sign-in is refused no longer gets stuck in a refresh loop.
    • Handover and ticket display — Handover notices show correctly after a page reload, and long ticket references wrap instead of overflowing.
    • Deleted knowledge stays deleted — Deleting a knowledge base while it is syncing can no longer bring its items back.
    • Declined cards — When your bank declines a card, the Console now says so in plain words.
    • Security hardening — Tighter link handling in the chat widget and stricter rules for the website crawler and visitor sign-in.
  5. v4.1.2

    This update fixes a problem that stopped some of Gydr's emails from reaching you. Console notification emails, account alerts and support ticket updates are now delivered again. Your in-app notifications were not affected.

    1 fixed

    Fixed1

    • Notification and alert emails — Some emails Gydr sends you were not being delivered. These included Console notification emails (for example, when you are added to an account), account alerts such as low credits or a channel that needs reconnecting, support ticket updates, and the test email from Settings → Notifications. They are now delivered as expected. Your Console notifications inbox kept working throughout.
  6. v4.1.1

    Console MCP is now included in every Gydr plan, Free included. You can connect Claude, Codex or another MCP client to your Console and let it manage your account. You pay only the small number of credits each request uses.

    1 changed

    Changed1

    • Console MCP on every plan — Connecting an AI assistant to your Console is no longer limited to Growth and above. Every account you belong to can be ticked when you connect an app, or added later under Settings → Connected Apps. Each request still uses credits from the account it runs in. If an account runs out of credits or has an overdue payment, its requests are refused before anything is sent.
  7. v4.1.0

    You can now manage your Gydr account from the AI assistant you already use — Claude, Codex or any MCP client — with exactly the permissions you have in Console. Your chatbot is more careful about saying "I don't have that" instead of stretching an unrelated answer, Design Studio previews every setting you change, and documents built from large scanned PDFs, tables and Chinese, Japanese or Korean text now come through whole.

    4 new6 changed7 fixed1 breaking

    Breaking Changes

    These may need a change on your side.

    • Event deliveries identify as GydrBot — Webhooks, Custom Desk calls and Conversation API callbacks now arrive with the User-Agent GydrBot/1.0 (+https://docs.gydr.ai/api/gydrbot) and signature headers. If your server or firewall allows requests by User-Agent, add this one.

    What's New4

    • Console MCP: manage Gydr from your AI assistant

      Connect Claude, Codex or another MCP client to your Console and ask it to do the work: add FAQs, products and locations, write and test a skill, build a widget and its design, read conversations and see why a reply came out the way it did, then fix it and test again. It acts as you, with your own Console role, and never sees your passwords or integration secrets. Anything that changes what visitors see, or deletes data, asks you first. Available on Growth and above; each request uses a small number of credits.

    • Choose which accounts each connected app may use

      When you connect an app you tick the accounts it may work in, and each app keeps its own current account — switching in one never moves another. Settings → Connected Apps lists every app with when it was last used and which account it is working in, and lets you change its accounts or disconnect it.

    • A help desk's API key lives on its integration

      For help desks that call Gydr with an API key, the key now has its own tab on the integration, with the address the desk calls and the desk's setup steps. Setting up the desk finishes in one place, right after you create it.

    • Signed event deliveries

      Webhooks, Custom Desk calls and Conversation API callbacks are now signed so firewalls and bot protection (for example Cloudflare) can recognise Gydr as a verified bot — no fixed IP address to allow-list.

    Changed6

    • New sign-in and sign-up screens — Signing in, registering and resetting your password now happen beside a short tour of what Gydr does, including a live demo chat in English, Malay and Chinese.
    • More honest answers — When nothing in your knowledge base answers a question, your chatbot now says so plainly instead of offering an unrelated FAQ or products the visitor never asked for.
    • Cleaner product recommendations — Each recommended product is described once, and only the products the reply actually names appear with it.
    • Turn insight shows what happened to your attributes — When a reply used a filter on one of your custom attributes, the insight now shows whether the filter applied and which items it left out.
    • Clearer import results — Product, FAQ and location imports report a readable summary such as "16 products created, 2 updated".
    • A redesigned developer documentation site — Easier to find your way, with every section laid out as a route you can follow.

    Fixed7

    • Large scanned PDFs — Image-only PDFs over 10 MB are now read with text recognition instead of failing with "no chunks".
    • Tables and Asian scripts in PDFs — Number-only table cells, and Chinese, Japanese or Korean text in some fonts, are no longer dropped from documents.
    • Testing a draft with a new knowledge base — A test turn on an unpublished draft now searches the knowledge bases attached in that draft.
    • Design Studio previews every control — Launcher position and icon, hover styles, the product panel side, the waiting line and text styles now show live on the canvas; dark-mode colours read correctly in the Style panel.
    • Wrong or contradicting product details — Your chatbot no longer states an attribute that the product's own value contradicts, or lists an item at a price your catalog doesn't carry.
    • Features during an overdue payment — Console no longer shows features as available while an account, or its partner, has an overdue payment.
    • Dashboard and conversation details — The active-widget count now follows the widget filter, and a conversation's first-message time is recorded.
  8. v4.0.0

    Major release

    Telling your chatbot who is signed in now takes one token: your page hands it over, Gydr checks it, and that person's history and memory follow them on every device. Handing a conversation to your team is now yours to shape — live chat, a ticket your team answers later, or both — with Zendesk and your own help desk joining the list. A new Design studio lets you style the chat widget by eye, and a Topics dashboard shows what your visitors actually ask about. This is a major release: a few old calls were removed, and a migration page lists each one with its replacement.

    17 new11 changed5 fixed10 breaking

    Breaking Changes

    These may need a change on your side.

    • WordPress plugin 1.4.x or older — visitors are guests until the plugin updates to 1.5.0.
    • Tokens your server signs for Gydr with a shared secret — sign-in stops and the old key is disabled. Use Your keys (your own key pair) or Your backend.
    • identify(), the old setters and the old data attributes — each logs one notice and does nothing, so details and sign-in are off until you move to setVisitorDetails, setVisitorTokenProvider or data-visitor-token-endpoint.
    • Visitor ids you set yourself with data-visitor-id — those visitors get a new id and history under the old id is not reachable. Add a sign-in source for cross-device history.
    • Conversation API calls with your own visitorId, identity or visitorContext are refused. Send userId or the returned visitorId, and visitorDetails.
    • The old visitor webhook event is gone — subscribe to visitor.details_updated. Custom Desk's verifiedSubject is gone — read visitor.userId and visitor.signInSourceId.
    • Desk links that used a custom id claim are switched off — re-enable them after choosing the sign-in source's User ID claim.
    • A visitor's End now closes the conversation in your help desk by default — switch off "Also close it" on the integration to keep it open.
    • Help desks now receive a private id for messaging-app visitors — turn on "Share the visitor's messaging-app user ID" if your desk matches visitors by their app user id.
    • Note for operators: once deployed, this release should not be rolled back on its own — doing so can briefly sign Console users out until the rollback is complete. Roll back together with the deployment layer that mints Console sessions.

    What's New17

    • One visitor token

      Give the widget the token your site already issues for the signed-in visitor, and choose in Settings → Visitor sign-in how Gydr checks it: WordPress (your site signs it with its own key, no code), Your keys (your identity provider's public keys, or one you paste) or Your backend (Gydr asks your server with a signed request). Plugins that call your API, MCP server, Shopify or WooCommerce as the visitor receive that same token, only on the address it belongs to.

    • A User ID for every signed-in visitor

      Signed-in visitors carry your own id for them. It shows on the visitor's row in the Console, in webhooks, the Conversation API, Custom Desk and exports, and you can search conversations by it.

    • History and memory that follow the person

      A signed-in visitor sees their conversations on every device, and their visitor memory is on automatically unless you turned conversation history off. Signing in, signing out or switching person always starts a fresh visitor, and a guest's conversation never carries into a sign-in. If your sign-in check is briefly unavailable, nobody is signed out.

    • Token tester and clear reasons

      Test a token in Settings → Visitor sign-in: it runs the real check and shows the User ID, or the reason the visitor was not recognised.

    • Conversation API: userId

      Create a session with the signed-in person's userId and they get the same visitor and history as on your website.

    • WordPress plugin 1.5.0

      The plugin creates its own signing key on your site, and order answers use the same token. It arrives through the normal WordPress updater right after this release.

    • Live chat, a ticket, or both

      Each help desk can now work in one of three ways. With live chat, a visitor talks to your team in real time inside the same chat window. With tickets, the visitor leaves their question and contact details, and your team replies later from your help desk — useful outside business hours or when nobody is online. Turn both on and the visitor chooses. For tickets, Gydr writes a short summary of the conversation so your team doesn't have to read the whole chat, and makes sure each ticket is filed only once. Filed tickets and their summaries use credits.

    • Zendesk, for every account

      Connect Zendesk for live chat, tickets, or both. Signed-in visitors are linked to their own Zendesk user, so your agents see who they are talking to.

    • Custom Desk

      Already use a help desk we don't connect to yet? Custom Desk lets your developers connect it through a documented, secure contract. It supports live chat, tickets, and letting your agents take over a conversation from the chatbot and hand it back.

    • Offer a person when visitors get stuck

      The chatbot now notices when a conversation isn't going well — the visitor sounds frustrated, repeats the same question, or goes several turns without a useful answer — and offers to bring in your team. The visitor can accept or keep chatting. It is on by default when a help desk is connected, and you can switch it off per widget.

    • Visitors can end a live chat

      Visitors can now leave a live chat with an End option instead of just closing the window. By default this also closes the conversation in your help desk, so your agents' queues stay tidy. You can turn either behaviour off per integration, and you choose this when you first set the integration up.

    • Conversation Topics

      See what your visitors ask about most, what they were trying to do, and whether they got what they needed. Gydr groups conversations into topics automatically, and you can add, rename or merge topics yourself. Results appear on your dashboard and can be exported. Topic analysis is opt-in and uses credits.

    • Design studio

      Style your chat widget on a live preview, no code needed: switch on dark mode, choose from curated fonts, add your logo to the header, set the launcher icon, and pick a Spotlight welcome screen or card-style suggested questions. You can preview the widget's form, handoff and offline screens before you publish. For more control, add your own CSS — Gydr checks it before saving and points to anything it can't accept. Advanced design options are on Growth and above.

    • Follow-up suggestions

      After each reply, the chatbot can suggest a few next questions as tappable chips, helping visitors keep the conversation going. Available on your website and on Messenger, Instagram and Telegram. Switch it on per widget; it uses credits only when suggestions are shown.

    • Notifications inbox

      Console notifications now live in an email-style inbox: see what's new at a glance, open a notification for the details, and mark it read.

    • Analytics and Handoff dashboards, now with exports

      Both dashboards now open with a short summary of how things are going, followed by the numbers behind it. Export the data to share with your team.

    • Rebuilt website import

      Importing your website into a knowledge base is simpler: choose from three clear import methods, pick exactly which pages to bring in, and see an estimate of the credits from your site's own pages before you start. The import stops on its own before it would use more credits than you have.

    Changed11

    • Gydr always makes the visitor id — a page, a link or an API call can no longer choose one, and returning browsers keep the id they already have.
    • Visitor details are called visitorDetails everywhere: the Conversation API, webhooks, Custom Desk and exports.
    • Desk links to Zendesk, Freshchat and LiveChat use a switch and a sign-in source, and fill the vendor's customer id from the User ID.
    • Memory is stored per account, widget and visitor.
    • Every knowledge base view shows which items your chatbot can actually read.
    • Every feature that uses credits now says so where you switch it on.
    • Saving visitor memory now uses credits, like other AI steps.
    • When your account is out of credits, the chat widget now says the assistant is unavailable instead of replying with a refusal.
    • Every fixed line a visitor reads in the chat widget is now translated.
    • A help desk that needs a visitor's name and email now asks for them before connecting.
    • Help desks now receive a private id for messaging-app visitors instead of their app user id. A switch per integration shares the real id.

    Fixed5

    • Closed several ways to read or change another visitor's data using only the public widget key — guessing an id to read details, history or memory, overwriting details or form answers without proof, and reaching a messaging-channel visitor from the website widget.
    • The Conversation API no longer returns an internal database id as the visitor id.
    • A website import now counts a redirected page once and crawls only the pages you selected.
    • Past conversations in the widget are labelled by your visitor's own calendar day.
    • Erasing a visitor's data now also ends their open chats, and the Console shows the erased state.

Gydr 3

24 releasesAugust 25 – October 7, 2026

  1. v3.7.2

    Your chatbot's welcome screen can now offer suggested questions that visitors tap to start a conversation, and you can arrange the welcome screen with a live preview. This release also fixes installing the WooCommerce plugin when you authorize it from your store.

    2 new1 fixed

    What's New2

    • Suggested questions on the welcome screen

      Add up to five questions your visitors can tap to start a chat — each is sent exactly as if they had typed it, so they get an answer straight away. Give the list your own heading, or keep the default.

    • Welcome builder with live preview

      Edit your welcome screen in one place and see the result as you type. Drag your featured FAQs and suggested questions into the order you want, on desktop or touch screens.

    Fixed1

    • Installing WooCommerce with "Authorize in WooCommerce" — finishing the install after approving Gydr on your store could fail with "Failed to save plugin configuration." The install now completes as expected; you can simply run the setup again.
  2. v3.7.1

    A small reliability release for the WordPress and WooCommerce connections introduced in 3.7.0. Reconnecting a WordPress site can no longer leave your signed-in visitors anonymous partway through, and authorizing your WooCommerce store now applies cleanly even if it is saved twice.

    2 fixed

    Fixed2

    • Reconnecting WordPress keeps your visitors signed in — approving a reconnect used to switch off your site's existing connection straight away. If the reconnect was never finished (the tab was closed or the code expired), your logged-in visitors chatted as anonymous until you connected again. Your existing connection now keeps working until the site has actually picked up the new one.
    • WooCommerce authorization applies once — saving the same store authorization twice at the same moment could apply it twice. A second save is now turned away with a clear message, and a save that fails can simply be tried again.
  3. v3.7.0

    Getting signed-in visitors working on your site no longer means editing code or copying keys between screens. Install the new Gydr Widget - AI Chat plugin on any WordPress site, press Connect, approve it in your dashboard, and you're done: the chat appears on your site and recognises your logged-in users. On a WooCommerce store, order lookups are one more approval, given inside your own store. Settings now gathers every way your visitors sign in into one list, and you choose which of them each plugin trusts.

    4 new4 changed8 fixed

    What's New4

    • The Gydr Widget - AI Chat plugin for WordPress

      A ready-made WordPress plugin that adds your chatbot to your site and tells it which logged-in user is chatting. Download it from the new Resources page, upload it to WordPress, and press Connect WordPress. You approve the site in your dashboard and the connection completes on its own. There is nothing to copy or paste, and no secret ever passes through your browser. The plugin works on any WordPress site, updates itself, and keeps your branding current, including White Label.

    • Order lookups switched on from your store

      When the site is a WooCommerce store, the same approval can let the assistant look up orders. Choose "Authorize in WooCommerce", approve on your store's own screen, and your store issues the assistant a read-only key on the spot. You never handle the key yourself. Pasting a key by hand still works if your store can't be reached.

    • Sign-in sources

      Settings → Signed-in visitors lists every way your visitors sign in: a WordPress site, an identity provider you already run, or your own site with a signing key. Add one from a single panel, wherever you need it: Settings, a plugin's setup, or Resources. Each plugin now trusts only the sources you tick, so one site's sign-in can't vouch for another store's customers.

    • Resources

      A new Resources page, under your avatar menu, holds the downloads you install on your own platform, starting with the WordPress plugin, with its version, checksum and changes.

    Changed4

    • Your existing signing keys carry over — keys created in the previous release appear as sign-in sources and keep working without changes on your site.
    • Simpler account menu — Resources sits beside Settings and Support, and every item has an icon. To sign out, go to Settings → Profile.
    • Partner roster shows what you allocated — the roster's "Monthly allocated" and "Pack allocated" columns show the credits you gave each client, are visible on every screen size, and update the moment you allocate.
    • Uninstall tells you what stays behind — removing a store plugin, or disconnecting a shop in WordPress, now tells you where to revoke the key your store issued, since we never change anything in your store.

    Fixed8

    • "Where is my order?" always looks the order up — the chatbot no longer answers a signed-in shopper with a form asking for their order number.
    • Order conversations stay private — a conversation that read a shopper's order is no longer listed in their past conversations or used for memory.
    • A clear prompt above the order form — guests now see a proper sentence above the order-number form instead of a bare field label.
    • The WooCommerce approval tab opens reliably — browsers no longer block it as a popup.
    • Plugin readiness is accurate — a plugin whose trusted sign-in source was removed no longer shows as ready.
    • Uninstall waits for the answer — the dialog stays open until the plugin is removed and shows any failure inside it.
    • Partner allocations are never charged twice — a temporary error while allocating credits can no longer lead a retry to deduct them a second time.
    • More time to finish WooCommerce setup — after you approve in your store, you have an hour to finish the rest of setup before you'd need to authorize again.
  4. v3.6.0

    Your chatbot can now know who it is talking to. Sign your logged-in users in to the chat from your own site, and the chatbot can do things only that person should see, starting with "where is my order?" on a WooCommerce store. Setting up WooCommerce is now a guided, step-by-step wizard that tests the connection as you go. And product answers are tighter: every product the chatbot names comes with its price, link and card, and what it says about a product comes from your catalog, not from guesswork.

    3 new9 changed5 fixed

    What's New3

    • Verified visitor identity

      Your website can tell the chatbot which signed-in user is chatting, with a token your server signs. Create a signing key under Settings → Identity verification, copy the ready-made snippet for your stack, and test a token right there before you go live. A verified visitor stays verified for the whole session, and a conversation that read their personal data is never shown to anyone else. The Integration API gains the same capability through a new identity field and identity:assert scope.

    • Order lookup for WooCommerce

      Opt in and your chatbot answers "where is my order?" A signed-in shopper sees their own recent orders. Anyone else is asked for the order number and the billing email, and only an exact match is answered. Order details never include the address, phone number or email.

    • Guided WooCommerce setup

      Adding a WooCommerce store is now a step-by-step wizard: enter the store address, test the connection, choose which capabilities to turn on, test each one, and finish. Every step shows what it checked and what to fix if it did not pass.

    Changed9

    • Product replies read like a knowledgeable advisor — each product the chatbot recommends gets its name, price, a line on why it fits, and the link, in the same shape on every channel.
    • Descriptions come from your catalog — a short product description is shown as you wrote it; a long one is summarised; notes you add to a product are used as facts the chatbot can draw on.
    • Every product named gets its card — when a reply mentions a product, it carries its price and link and appears in the product panel, even when the search returned more products than the panel can show.
    • Stock status comes from your data — "In stock" or "Sold out" appears under the price only when your catalog or store says so. The chatbot no longer writes its own availability claim.
    • Sold-out products are still shown — a visitor who asks about a product you have marked unavailable is told so, instead of being told it does not exist.
    • Long documents are read in full — answers drawn from a long knowledge base document now use the whole relevant passage, so the chatbot no longer tells a visitor your business does not carry something that is in your document.
    • Product attributes are quoted — custom attributes on your products (origin, size, grade and the like) are used in replies and shown as facts.
    • Clarifying questions stand alone — the chatbot no longer answers a question and then asks a clarifying form underneath the same reply.
    • Privacy policy updated — the policy now describes identity verification and what the chatbot can see about a verified visitor.

    Fixed5

    • No invented sentences under products — the chatbot no longer writes its own line of description under a product whose catalog entry is only a name and price.
    • A bold question is a question — a reply that ends on a bolded question is no longer mistaken for a statement.
    • The chatbot never denies your catalog — a correction step could tell a visitor you do not carry something that your document lists. It cannot any more.
    • Order and recall answers are grounded — follow-up questions about an order, or about something said earlier in the chat, are answered from what the chatbot actually retrieved.
    • A visitor who signed in keeps every feature — forms, location answers and file uploads continue to work after the chatbot has recognised a signed-in visitor.
  5. v3.5.0

    Your chatbot can now sell from a WooCommerce store, and every connected store searches better: you choose how results are ranked, and a chatbot with several product sources checks all of them for every question. Product answers read more consistently, with price ranges for products that come in variants, and a new help desk integration joins human handoff.

    4 new8 changed5 fixed2 breaking

    Breaking Changes

    These may need a change on your side.

    • Shopper sign-in tools are switched off — Shopify tools that need a signed-in shopper (viewing a cart, listing or looking up orders) are turned off, because a chatbot cannot sign a shopper in. Product browsing and search are unaffected.
    • score removed from API product results — products[] in conversation responses and the message.completed webhook no longer includes score. The field was optional; integrations that ignored it need no change.

    What's New4

    • WooCommerce Storefront plugin

      Connect your WooCommerce store and your chatbot can browse, search and recommend your products, with prices, images and links straight from your catalog. It reads your public storefront only, so there are no keys to create or share.

    • Store search settings, including Smarter ranking

      Shopify and WooCommerce plugins gain a Search tab. Tell the chatbot which language(s) your catalog is written in, and turn on Smarter ranking to reorder store results by how well they match the visitor's question. Smarter ranking is optional and appears on its own line in your usage.

    • A new help desk integration for human handoff

      You can now hand conversations to one more help desk, with the agent availability check and two-way messaging.

    • Plugin categories, filter and search

      Plugins are grouped into categories with icons, and both plugin pages let you filter and search, so the plugin you need is quick to find.

    Changed8

    • Every product source is searched — a chatbot with a knowledge base and one or more stores now searches all of them for each product question and combines the results fairly.
    • Price ranges for products with variants — product cards and replies quote the full price range instead of the lowest price, including in your conversation transcripts.
    • Product cards follow the reply — the product panel shows items in the order the reply mentions them, and every product is presented the same way whichever source it came from.
    • Replies stay in your visitor's language — product descriptions written in another language are translated in the reply.
    • Links come from your catalog — the assistant only shares links your products and content actually contain.
    • A simpler availability check — your help desk's agent availability check now uses the integration's existing credentials. The separate access token introduced in v3.4.0 is no longer needed.
    • Clearer catalogs to add from — plugins, channels and integrations are listed as rows with a description and one action each. Adding and editing an integration now share the same tabs.
    • Honest store expectations — store plugin pages explain what a connected store can and cannot do, and the chatbot no longer offers a cart it cannot manage.

    Fixed5

    • Line breaks render as line breaks — replies no longer show a literal <br> inside tables and lists.
    • The availability Test button works again on the integration page.
    • A dismissed form stays dismissed — the assistant no longer asks the same form again after a visitor closes it.
    • Product descriptions appear on category searches — browsing a category no longer hides each product's description.
    • Malay and Indonesian are told apart by region when a message could be either.
  6. v3.4.0

    Your plugins now tell you when they are failing and when they have recovered, in the list, on the plugin page, in your notifications and by email. Team invitations are checked before they are sent, so a mistyped address is caught on the spot instead of vanishing. And every error the Console shows now says what went wrong and what to do about it.

    3 new6 changed6 fixed1 breaking

    Breaking Changes

    These may need a change on your side.

    • Enter your help desk availability access token — if your handoff integration has the agent-availability check enabled, it reports agents as unavailable until you paste the new access token into the integration's Human Handoff Fallback section and save. Until then, visitors are not offered a handoff for that desk.

    What's New3

    • Plugin health, reported where you look

      When a plugin stops working — its credentials are rejected, its endpoint cannot be reached, it returns errors, or it is being rate-limited — the plugin is marked Failing in your Plugins list, and its page states the cause and since when, with the repair action one click away. Everyone who can view Plugins gets a notification when a problem opens and another when it clears; you can also have a failure and recovery email sent to the addresses of your choice under Settings → Notifications. A brief hiccup does not raise the alarm: a transient failure has to repeat before it counts, and the first successful call closes the case.

    • Invitations are checked before they are sent

      Inviting a team member now verifies the address first. An address that does not exist, or that has already rejected our email, is refused right on the email field with a plain explanation, so a typo is fixed in seconds rather than discovered days later when nobody joined.

    • Human handoff checks agent availability with your own access token

      Your help desk integration's availability check now runs on an access token issued to your organisation. Enter it in the integration's Human Handoff Fallback section, press Test to confirm it works, and save.

    Changed6

    • Every error tells you what happened — messages across the Console now name the specific problem and the next step, in your language, and land on the field they concern. Generic “something went wrong” wording appears only when there is genuinely nothing more to say, and technical codes are shown only where they help a support conversation.
    • Product questions always consult your store — a chatbot whose products come from a connected store now searches it for every product question, so a visitor asking what you sell gets an answer from the catalog rather than a guess.
    • A search that finds nothing is not “we don't sell that” — when a store search returns no results, the assistant tries another wording once and, if still empty, says it could not find a match and invites the visitor to rephrase. It no longer claims your store does not carry something.
    • Plugin and channel pages share one tab layout — configuration pages use the same horizontal tabs at every screen size.
    • Google Sheets structure is shown, not edited — the Data Logging tab now presents your sheet names and column headers as read-only information refreshed from the spreadsheet, since the spreadsheet is the source of truth.
    • A channel credential alert you can subscribe to — Settings → Notifications gains a row so you can be emailed when a connected channel's credentials stop working.

    Fixed6

    • The product panel shows up for store-only chatbots — a chatbot whose only products come from a connected store now displays product cards to visitors; previously the cards were produced but the panel never appeared.
    • Resending an invitation no longer disappears silently — if an address had bounced, resend and reactivate reported success while nothing was delivered. Both now explain that the address cannot be reached and what to do instead.
    • Store-only chatbots no longer mention a knowledge base they do not have when declining to give a catalog count.
    • Shopify setup guidance names the right permission — the install instructions asked for an access scope Shopify does not offer.
    • A bad gateway response no longer shows raw parser text — you see a clear server-error message and can retry.
    • Google Sheets quota errors say to wait, not to re-share the spreadsheet.
  7. v3.3.4

    Google Sheets connections now run on a service account you own, so a connection no longer stops working on a timer you were never told about. The widget editor has been reorganised into ten clearly named tabs, and the Plugins pages have been rebuilt to match the rest of the Console.

    1 new8 changed3 fixed1 breaking

    Breaking Changes

    These may need a change on your side.

    • Existing Google Sheets connections must be reconnected — the previous sign-in method is retired. Open each Google Sheets plugin, follow the walkthrough to create a service account, and paste its key file. Your sheet tabs and column mapping are kept; until you reconnect, the plugin reports that the connection needs to be redone rather than working.

    What's New1

    • Connect Google Sheets with your own service account

      Create a service account in your Google Cloud project, share the spreadsheet with its email address, and paste the downloaded key file once. Setup checks the connection on the spot and tells you exactly what to fix when it fails — the spreadsheet isn't shared, the ID is wrong, or the Sheets API isn't enabled in your project. A step-by-step walkthrough guides you through it, and you can replace or remove the account later from the plugin's Connection tab without losing your sheet and column mapping.

    Changed8

    • The widget editor has ten focused tabs — General, LLM Configuration, Skills, Conversation, Visitor, Data, Integrations, Channels, Security and Settings, ordered by how often each is edited. Visitor context, memory and privacy have their own Visitor tab; Data Forms have their own Data tab; allowed domains, IP ranges and HTTPS live together under Security; and Channels lists your Hosted Widget, SDK (with its API keys) and external channels under the same names the Connect dialog uses.
    • Plugins are easier to find and attach — the Plugins list now shows each plugin's type, status and description with search and card/list views, and attaching a plugin to a skill works the same way as attaching a channel to a widget.
    • Google Sheets errors say what actually went wrong — a sheet shared as Viewer, an unshared or mistyped spreadsheet, a disabled Sheets API and a revoked key are each reported as their own cause, with the fix named.
    • Long pasted values keep their shape — plugin fields that expect a multi-line value, such as a key file, now accept it without collapsing it into one line.
    • Your test page copes with long answers — a test turn that takes a while no longer risks timing out before the answer arrives.
    • MCP connectors work with more servers — servers that still use the older streaming transport now connect, and sessions are reused between calls.
    • Shopify carts persist through the session — a visitor's cart carries across turns instead of starting over.
    • A visitor's data is erased in one action — the erase request removes the visitor's conversations, memory, form responses, attachments and reports together.

    Fixed3

    • Test page citations now show their content — the debug panel listed every citation as empty; it now shows the source, its relevance and the matched excerpt.
    • A retried form submission is no longer counted twice — if the first page of a Data Form is submitted again after a lost response, you are billed for one response, not two.
    • Viewers no longer see a blank editor — a team member with view-only access opening a widget that has never been saved now sees a clear empty state.
  8. v3.3.3

    WhatsApp now lets people message a business without sharing their phone number. Your assistant understands those messages and replies to them, on every WhatsApp number you have connected.

    1 fixed

    Fixed1

    • Messages from senders without a shared phone number now arrive and get a reply — WhatsApp identifies such senders by a business-scoped ID instead of a phone number. Your assistant now recognises them, keeps their conversation together across visits, and replies to them like any other visitor.
  9. v3.3.2

    A small follow-up to v3.3.1. Messages sent to a WhatsApp number you connected from the WhatsApp Business app could still be dropped for some senders. Your assistant now receives them.

    1 fixed

    Fixed1

    • WhatsApp messages from every sender now arrive — a message whose sender identity was delivered in a newer format could be dropped before your assistant saw it. It is now understood and answered like any other message.
  10. v3.3.1

    This release makes your Shopify storefront a first-class source for your assistant, so questions about store products are answered from what the store actually holds and shown as product cards. It also fixes a handful of ways your assistant could over- or under-answer a catalog question, and makes sure messages sent to a WhatsApp number you connected from the WhatsApp Business app reach your assistant.

    3 changed5 fixed

    Changed3

    • Shopify products answer like catalog products — when your assistant finds items in your Shopify storefront, it presents them the same way it presents your own catalog: grounded in the store's data and shown as product cards.
    • An honest “I don't know” starts with what you do have — when a visitor asks about a detail your catalog doesn't record for a product, the reply now opens with the product itself before saying which detail is missing.
    • Stock is shown as available or not — Shopify searches no longer ask for an exact stock count, so they work for every store regardless of the access it granted.

    Fixed5

    • “How many products do you have?” gets a number — your assistant now answers catalog-count questions with the actual total instead of saying it couldn't find what you were looking for.
    • No invented descriptions — for an item whose only details are a name and a price, your assistant shows exactly that and no longer fills in ingredients, benefits or other details it was never given.
    • No product pitch on top of a caution — when your assistant advises a visitor to consult a professional, it no longer places a product card above that advice.
    • WhatsApp messages on a Business-app number now arrive — a number connected by keeping it on your WhatsApp Business app could have its incoming messages silently dropped. They now reach your assistant.
    • Shopify searches no longer fail on stores without inventory access — a store that hadn't granted stock-level access returned no products at all.
  11. v3.3.0

    Your assistant can now gather structured information from visitors and hand it back to you as a table you can export. Data Forms let you author the questions, attach them to a widget, and have them asked naturally — in the visitor's own language, on your website and every connected channel. This release also makes reconnecting a channel painless, adds Google Sheets workbook loading, and sharpens how the assistant answers questions about your catalog.

    6 new4 changed3 fixed1 breaking

    Breaking Changes

    These may need a change on your side.

    • Note for operators: this release changes what a Console sign-in token can carry. Once deployed, it should not be rolled back on its own — doing so can briefly sign Console users out until the rollback is complete. If a rollback is ever needed, roll it back together with the deployment layer that mints Console sessions.

    What's New6

    • Data Forms — collect answers inside the chat

      Author your own forms under Data → Forms, attach one to a widget, and your assistant asks the questions at the start of a session or when the conversation calls for it. Answers collect in a table you can browse, filter and export as CSV. They're asked in whatever language the visitor is writing in, with no per-language setup; contact details you already hold are filled in for you, so a returning visitor isn't asked for their name and email again; and it works on your website widget and every connected channel. Each response collected costs one credit, counted once per response however many questions it spans, and responses are kept for 12 months from when they're first received.

    • Record answers in one language of your choosing

      Free-text answers can optionally be stored in a single language for your team to read, with the visitor's original wording kept alongside it.

    • Reconnect a channel without hunting for credentials

      When a channel's connection expires, reconnect it from Console by re-running the same guided sign-in. Nothing to find, nothing to paste.

    • Invite a team member straight onto a custom permission set

      Pick the permission set as part of the invite, instead of inviting first and adjusting their access afterwards.

    • Google Sheets: choose a sheet instead of typing its name

      Connect a workbook and Gydr loads its sheets and column headers for you to pick from. You can also give each connected sheet its own instruction describing when and how your assistant should write to it.

    • Erase a visitor's data in one action

      From the visitor view, remove everything a visitor gave you — their remembered details and every Data Form response across every form — in one confirmed action.

    Changed4

    • More accurate answers about your catalog — when an item has no description, your assistant now gives its name and price only, rather than filling in details it doesn't have.
    • A better read on what a visitor is asking — mentioning a personal circumstance mid-conversation, like an allergy or a budget, now gets an answer instead of switching into product recommendations.
    • Product cards appear when you'd expect them — when a visitor asks to see or buy something, not when they ask a general question.
    • Clearer Data Form editing — Console now says exactly what a refused save was refused for, and mentions the question-key freeze once rather than on every question.

    Fixed3

    • First answers on a new form no longer fail — submitting the first answer could return an error; it now saves correctly.
    • Multi-page forms continue after contact details — submitting your contact details no longer dropped the next page of a form.
    • A first-time visitor's answer counts — an answer given before a conversation has started is now recognised by a session-start form.
  12. v3.2.9

    When you report a conversation that went wrong, you can now see it, change your mind about it, and take it back. We've also fixed a problem where asking the chatbot for your nearest location could leave the conversation stuck, and tidied up the conversation screen on phones.

    2 new1 changed2 fixed

    What's New2

    • See and manage the issues you've reported

      Reporting a conversation now gives you a confirmation you can come back to, and the report stays visible from the conversation you filed it against. You can find it again without hunting for it.

    • Change your report, or take it back

      While a report is still waiting to be looked at, you can revise what you wrote — the category, your comment, and your notes on individual replies. And for as long as it's open, you can withdraw it entirely. A withdrawn report is closed for good: nobody on our side can reopen it, work on it, or read it as though you were still asking.

    Changed1

    • One report per conversation — if you already have an open report on a conversation, filing again simply opens the one you have instead of creating a duplicate. Previously it was easy to file the same issue twice without realizing.

    Fixed2

    • The chatbot no longer gets stuck asking for your location — when you asked for your nearest store or branch, the chatbot could show you a location box that it then refused to accept, telling you the form was no longer active and asking again, sometimes several times in a row, with no way out except ending the chat. The location box now works the first time.
    • The conversation screen is readable on a phone again — reading a conversation on a small screen had become cramped: the date broke across several lines, the visitor's name was pushed off the edge, and the buttons ran past the side of the screen. Report and Export now sit on their own row, and everything else has room to read.
  13. v3.2.8

    When your assistant gets something wrong, you can now tell us from inside the Console — pick the replies that were off, say what happened, and send. Everything we need to investigate goes with it automatically. This release also stops your chatbot quoting a price or a phone number it cannot support.

    2 new2 fixed

    What's New2

    • Report a conversation

      In Visitors, open a conversation and choose Report an issue. Tick the replies that were wrong, add a note against any of them, pick what went wrong and send. You'll see the report's progress on the conversation, and we'll let you know when we've looked into it and how it was resolved.

    • Erase a report

      Reports keep a copy of the conversation so we can investigate. If a visitor asks you to erase their data, you can now remove a report and its stored copy from the conversation view.

    Fixed2

    • Invented prices and phone numbers — a reply could state a price or a hotline number that appears nowhere in your knowledge base. Your chatbot now checks any figure it is about to quote and won't state one it cannot support.
    • A blank screen when switching accounts — the Console could fail to load after signing in or moving between accounts.
  14. v3.2.7

    A small reliability release. When we shipped an update, there was a brief window in which the Console could sign you out, and signing back in would drop you again until the update finished. This release closes that window, so updates pass without interrupting your session.

    1 fixed

    Fixed1

    • Staying signed in during updates — while an update was rolling out, the Console could sign you out and refuse the next sign-in until it completed. Your session now carries through updates without interruption.
  15. v3.2.6

    When your chatbot double-checks one of its own answers, it now re-reads the information it already found instead of running a new search. This release also closes several branch-lookup gaps and keeps replies in the language the visitor is writing in.

    2 changed5 fixed

    Changed2

    • Branch questions with a named city — asking about a city where you have no outlet now gets your nearest outlets, not a request for the visitor's location.
    • "Is there another branch?" with a known location — the chatbot now looks further afield (up to 100 km) before saying nothing is nearby, keeps any branch-type filter the visitor asked for, and says so plainly when none is within range.

    Fixed5

    • A policy question could get a "couldn't find that" reply — when the chatbot re-checked an answer, its second attempt could end with nothing to say. It now rewrites from what it already found, so the answer is given.
    • Branch lists no longer appear during a re-check — the re-check step can no longer run a store search of its own, which is what produced the unrequested outlet lists.
    • The location form no longer re-appears right after it is submitted.
    • Replies stay in the visitor's language — a foreign brand, wine or place name inside an answer no longer causes the reply to be rewritten, and a conversation in one language no longer switches to a closely related one (for example Italian to Malay) on a shared word.
    • Privacy in the re-check step — a visitor's own contact details are never used as reference material when a reply is re-checked.
  16. v3.2.5

    A follow-up to v3.2.2. Two remaining cases where a branch question could be answered with a list of outlets before your chatbot knew where the visitor was are closed.

    2 fixed

    Fixed2

    • A branch question with a store-type hint still skipped the location ask — when the search narrowed to "stores" or "clinics" but the visitor had named no place, outlets from an arbitrary area could still be listed. Narrowing by type no longer counts as knowing where the visitor is; the location form is shown first.
    • A location the visitor typed long ago was reused without asking — outlets near a previously typed location could be listed on a new visit. A typed location is now confirmed once via the prefilled form before it is used; a location supplied by your site or app is still used directly.
  17. v3.2.4

    Your chatbot now recognises thirteen more languages when a visitor writes in them, and it tells Malay and Indonesian apart reliably, so each visitor is answered in the language they actually wrote.

    1 new1 changed

    What's New1

    • Thirteen more languages recognised

      A visitor who writes in Italian, Dutch, Polish, Turkish, Czech, Danish, Finnish, Hungarian, Norwegian, Romanian, Swedish, Swahili or Tagalog is now answered in that language. Until now those messages were answered in English.

    Changed1

    • Malay and Indonesian are told apart reliably — the two languages share many words, and a Malay question could be answered in Indonesian or the reverse. Your chatbot now reads the words that only one of them uses and replies in the right one. A question that is genuinely identical in both languages follows the language the conversation is already in.
  18. v3.2.3

    A patch on top of v3.2.2. A visitor who writes in one language and mentions a place, product, or brand name from another could get their reply in that other language. Your chatbot now stays with the language the visitor is actually writing in.

    1 fixed

    Fixed1

    • Replies switching language on a place name — a visitor chatting in English who typed something like "I'm at around Kepong" could receive the whole reply in Malay, because the town name outweighed the sentence around it. A reply now changes language only when the visitor's message is actually written in that language. A genuine switch still takes effect immediately, in every direction.
  19. v3.2.2

    A patch on top of v3.2.1 for businesses with more than one outlet. When a visitor asks about a branch without saying where they are, your chatbot now asks for their location first instead of recommending an outlet at random.

    1 fixed

    Fixed1

    • Branch suggestions for visitors who haven't said where they are — a question like "can I exchange this at another branch?" or "which outlet should I go to?" with no place named could be answered with a store in the wrong city. Your chatbot now shows the location form (or asks for a city on messaging channels) and only names outlets once it knows where the visitor is. A visitor who has already shared a location gets the outlets nearest to it, and a visitor who names a city gets that city's outlets as before.
  20. v3.2.1

    A small patch on top of v3.2.0: Instagram direct messages sent to a connected Instagram account were received but never answered. They are now.

    1 fixed

    Fixed1

    • Instagram replies — a message sent to your connected Instagram account was accepted but dropped before your chatbot saw it, so nothing came back. Instagram conversations now get a reply like Messenger and WhatsApp.
  21. v3.2.0

    This release gives you finer control over who can do what in your Console. Build custom permission sets that grant exactly the access a teammate needs, review everyone's access at a glance, and get notified in the Console itself when something needs your attention. Permission changes now take effect instantly, with no sign-out required. You can also connect a WhatsApp number that stays on your phone, and connecting Facebook Messenger and Instagram from the Console works again.

    5 new8 changed6 fixed

    What's New5

    • Custom Permission Sets

      Owners can now create permission sets that grant exactly the pages and actions a teammate needs, beyond the four built-in roles. Assign a set to any Member or Viewer seat, change it later, and every holder's access updates immediately.

    • A Read-Only Viewer Role

      New invitations now default to Viewer, a read-only seat that can browse analytics, conversations, widgets, skills, knowledge, and integrations without changing anything. Pages a Viewer cannot use are hidden rather than greyed out.

    • Team Access Review

      Settings → Team now explains what each role and permission set can do, and lets you export a review of every seat's access for your own records or an audit.

    • Console Notifications

      A new bell in the Console collects notifications addressed to you: low or exhausted credits, billing events, support replies, and team changes. Each person controls which of these also arrive by email, and only sees notifications for the areas their access covers.

    • Keep Your WhatsApp Number on Your Phone

      Connect a number you already use in the WhatsApp Business app without giving it up. Customers keep reaching you on the phone, your chatbot answers from Gydr, and when a person replies from the phone the bot steps aside for that conversation. A banner in the Console shows who is holding it, and "Return to bot" hands it back.

    Changed8

    • Permission changes apply instantly — changing a teammate's role or set, removing them, or suspending an account takes effect on their very next request. Nobody is asked to sign in again.
    • Suspended accounts get a clear screen — a member of a suspended account can still sign in and sees an explanation, instead of a failed login.
    • Partner staff leave with the partner — when a partner's membership on a managed account ends, the seats it granted to partner staff end with it.
    • Faster sign-in and account switching — signing in and switching between accounts is noticeably quicker, and Console pages respond faster after idle periods.
    • Deleting a channel releases its Page or number — removing a Messenger, Instagram or WhatsApp channel lets that Page or number be connected elsewhere, unless another of your channels still uses it.
    • New WhatsApp numbers are ready to send immediately — a number onboarded through the guided WhatsApp setup is registered for you; no extra step before the first message.
    • No more automatic follow-up message on WhatsApp — Gydr no longer sends a templated nudge when a conversation goes quiet.
    • Opted-out WhatsApp recipients fail clearly — a message to someone who has blocked or opted out is marked failed straight away instead of waiting to be retried.

    Fixed6

    • Admins can manage Viewer seats — the row menu on a Viewer teammate was empty for Admins; it now offers the same actions the account allows.
    • Publish stays reachable on a saved draft — after saving a widget draft, the Publish action could disappear from the editor bar. It now stays available.
    • Facebook Messenger and Instagram connect again — the connect pop-up finished but the Console showed nothing, or reported that the authorisation could not be completed. Both flows now connect end to end.
    • Instagram replies on a Page-linked account — replies were rejected for some Instagram accounts linked to a Facebook Page; they now send.
    • Channel settings edits keep the rest of your settings — saving one field on a channel no longer clears the others.
    • Console header no longer overlaps page content on some screen sizes, and documentation links in the Console open the right page.
  22. v3.1.0

    This release opens up the Gydr partner program: partners get a full profile and a public directory listing, and prospects can now book a demo straight from our website — routed to the right partner automatically. Developer documentation has moved to its own dedicated site, and knowledge base documents gained a portable editing workflow.

    5 new2 changed6 fixed

    What's New5

    • Partner Profiles and a Public Partner Directory

      Partners can now manage their own identity in the Console — trading name, logo, website, and contact details — and opt in to a public partner directory on gydr.ai, where prospective customers can find them.

    • Book a Demo

      Prospects can request a demo directly from the website. Each request is verified by email and routed to a matching partner, who sees and responds to it from a new demo desk in the Console.

    • A Dedicated Developer Docs Site

      Developer documentation now lives at docs.gydr.ai, with faster navigation and a cleaner reading experience. All existing documentation links redirect automatically.

    • Duplicate Widgets and Skills in One Click

      Copy an existing widget or skill straight from its list to use as the starting point for a new one.

    • Editable Document Chunks

      Export a document knowledge base's content as a single editable file, refine it, and import it back — into the same knowledge base or a different one.

    Changed2

    • Simpler partner types — the partner program now has two types: Managed Service Partner and Technology Partner. Existing partners have been transitioned automatically; no action needed.
    • Clearer partner credit management — allocating credits to a managed account now shows a preview of the result, asks for confirmation before moving anything, and explains exactly why a transfer can't proceed.

    Fixed6

    • Form errors now name the field — a rejected save tells you which field was refused and why, instead of a generic “failed to save” message.
    • Chat attachments could silently fail — an upload processing error could discard a visitor's attachment without any sign of failure. Uploads now recover reliably.
    • Occasional undelivered replies — in rare cases a generated reply failed to reach the visitor's chat window. Delivery is now verified end to end.
    • Usage charts could under-report — daily usage on the credit consumption chart could show less than the monthly total under heavy traffic. The gap now closes automatically.
    • Partner credit reclaim is precisely bounded — a partner reclaiming credits from a managed account can only reclaim what they originally granted.
    • Integration setup no longer flashes the wrong form — a connect step now shows a loading indicator while its configuration is being fetched.
  23. v3.0.3

    This release sharpens how your assistant chooses the language it replies in. Conversations now stay in your visitor's own language reliably, and the full range of languages Gydr supports — including those written in non-Latin scripts — is handled consistently from the first message to the last.

    3 fixed

    Fixed3

    • Replies stay in your visitor's language — in rare cases a conversation could receive an answer in a language nobody in the chat had used. Your assistant now consistently answers in the language your visitor is actually writing in.
    • Consistent quality across every language we support — answers in languages such as Hindi, Bengali, Tamil, Hebrew, Greek, Burmese, Khmer and Amharic are now treated exactly like every other supported language, so they arrive as written instead of occasionally being reworded or slipping into English.
    • Language settings on your knowledge base are always recognised — if you record your knowledge base's content language using a short two-letter code, that setting is now read correctly.
  24. v3.0.0

    Major release

    This is the largest release Gydr has shipped. Your assistant no longer lives only on your website: connect WhatsApp Business, Facebook Messenger, Instagram, or Telegram to a widget, and that same widget answers on every one of them — the same knowledge, the same skills, the same personality. Your website chat is simply one channel among them. Alongside that, the whole chat experience now speaks fifty languages, human handover works properly on every surface, and the Console screens you use daily have been rebuilt.

    17 new10 changed20 fixed

    What's New17

    • Four Messaging Platforms, One Assistant

      Connect WhatsApp Business, Facebook Messenger, Instagram, or Telegram from the new Channels page and attach it to a widget. That widget starts answering there immediately — nothing is configured twice, because a channel is a new way to reach the widget you already built, not a second assistant to maintain. TikTok is listed as coming soon.

    • Guided Connection

      For WhatsApp Business, Messenger, and Instagram you can connect by signing in with Facebook and choosing your account, instead of copying credentials by hand. Telegram connects with your bot token and registers itself. Every connect form opens with the specific steps to follow before you start, and your credentials are verified with the platform before the channel is saved — so a wrong token fails at connect, not silently at the first customer message.

    • Replies Shaped for Each Platform

      Each platform receives your assistant's answer in its own vocabulary — product lists, buttons, quick replies, images, and location details rendered the way that platform expects, within that platform's real limits. Answers are sized to fit, so a long reply is never silently cut in half.

    • Forms That Work in a Chat Thread

      When your assistant needs structured information on a messaging channel, it asks one question at a time in the conversation rather than sending a form the platform cannot display. Tapped buttons are recorded as real answers, and each answer is saved as it arrives, so a long question set is never lost partway through.

    • Attachments Both Ways

      Images and files a customer sends on a channel are collected promptly and attached to the conversation, and files your team sends back are delivered natively on the platform rather than as a bare link.

    • Every Conversation in One List

      Because every channel answers through the same widget, its conversations land in the same list as your website chats — filterable by platform, with the channel shown on each conversation. Your dashboard reports per-channel too, so you can see where customers actually reach you. And when a messaging platform closes a conversation after a period of silence, the thread carries forward when your customer returns rather than starting cold.

    • Human Handover on Messaging Channels

      When a customer on a messaging channel asks for a person, the conversation transfers to your team exactly as it does on your website — your agent's availability is checked first, the customer is told what is happening, and your assistant steps aside until your agent closes the conversation.

    • Agent Replies Are Held, Never Lost

      If your agent answers after the platform's messaging window has closed, the reply is held for up to 72 hours and delivered the moment the customer writes back. Your team is told when a reply is held, when it is delivered, and when it expires — and can nudge a quiet conversation back open with an approved follow-up message. Held-reply status shows per conversation and per channel in the Console. Your agent is also told when a reply is too long for the platform before it is sent, and a message that permanently fails to deliver is recorded in your account's audit trail.

    • Fifty Languages, Across the Whole Chat Window

      Buttons, form questions, progress messages, error notices, rating prompts, and system messages all now match your visitor's language across fifty of them — not only the assistant's answers. A visitor writing in Thai no longer meets English controls around a Thai conversation, and the language is recognised from their very first message.

    • Right-to-Left Support

      For Arabic, Hebrew, Persian, and Urdu the chat window mirrors, so the layout reads in the same direction as the text.

    • Channel Health Monitoring

      Every connected channel's credentials are checked daily. If a platform revokes or expires a connection, you are alerted, the channel's real platform status is shown in the Console beside its own status, and Instagram connections are refreshed automatically before they expire. Turns that a dead channel could never deliver are no longer billed.

    • Business Timezone

      Set your business timezone once in Account settings. Your usage charts, dashboards, and channel date and time handling all follow your calendar day instead of a fixed one — so "yesterday" means your yesterday.

    • A Rebuilt Dashboard and Usage History

      Both were redesigned as single instruments rather than a grid of tiles: clearer credit consumption over time, sharper conversation metrics, per-channel reporting, and far faster loading across every date range.

    • A Desk-Style Visitors Page

      A persistent information rail on desktop keeps a visitor's details beside their conversation, and the mobile experience was rebuilt around an identity strip and a bottom sheet. Channel icons show where each visitor came from.

    • Draft and Live at a Glance

      The widget editor collapses into one row showing what is published versus what you are editing, with save actions in a sticky bar that stays clear of the preview. Widget Studio is now a mode you toggle within the editor rather than a separate destination, so you can move between visual editing and settings without losing your place.

    • A Rebuilt Connect Modal

      Connect is now a list of destinations with a tab per messaging channel, so getting your widget onto a website or a platform is one place instead of several.

    • Larger API Documentation Imports

      The REST API plugin accepts OpenAPI documents up to 25 MB, uploaded directly, and generates its configuration in stages so large specifications complete reliably.

    Changed10

    • Consistent controls everywhere — every filter bar, action cluster, and button in the Console follows one layout grammar and one control height, on every screen size, with buttons sized correctly on mobile.
    • Usage in your own calendar — the credit chart buckets by your business day, so daily and monthly totals agree and a full day means a full day.
    • Usage History loads as one view — both of its data sources are read together instead of in sequence, ending the timeouts on wide date ranges.
    • Accurate dashboard figures — visitor counts in the language panel and handoff rates are now measured against one consistent population, and every series is bucketed in your own calendar.
    • Bounded dashboard ranges — the date range is capped to a span that returns promptly rather than appearing to hang.
    • Import errors that point at the problem — a rejected knowledge CSV names the row and column that failed.
    • Validation errors that name the field — a rejected form says which field and why, instead of "Failed to save."
    • Complete audit log pages — a filtered audit log fills the page instead of truncating it.
    • Channel deletion is clean — removing a channel detaches it from your live widget and from your saved configurations, rather than leaving a broken reference behind.
    • Configuration screens match — channels and plugins share one configuration layout and a real loading state, and a field offering a fixed set of choices renders as a choice rather than a masked secret.

    Fixed20

    • Asking for a person no longer ends the chat — a request for a human hands the conversation to your team and keeps it open; only a genuine goodbye closes it. Saying goodbye no longer opens a ticket, either.
    • A handover never reads as a hang-up — a conversation being transferred no longer says goodbye on its way to your agent.
    • You are always told what happened — every handover outcome is announced to the customer, and if it cannot go through they are given your alternative contacts rather than silence.
    • No promises we cannot keep — a widget with no help desk connected never offers to fetch a human.
    • The conversation belongs to your agent — once a person is handling it, nothing else can end it out from under them, and closing it on your side ends the session properly and tells the customer.
    • A customer message during handover always reaches someone — a message sent while transferring is never quietly dropped.
    • Similarly-named products — questions about one product in a family of near-identical names now resolve to the right one, or ask a short clarifying question instead of guessing.
    • Nearby-location answers — "what is near me" questions complete in one turn, name the right places, and the locations shown always match the ones named in the reply.
    • Product cards match the message — the cards below a reply never show more, fewer, or different products than the reply itself names.
    • Fewer irrelevant results — weak matches are no longer presented as though they were answers, and searches with filters applied still tolerate a reasonable misspelling.
    • The assistant only offers what it actually has — it no longer mentions capabilities that were withdrawn for that turn, and the response settings configured on a widget now reach the conversation.
    • No repeated buttons on messaging channels — a button from a finished conversation can no longer restart it, and tapping one no longer produces a fresh copy of itself.
    • Taps are shown and understood — a tapped option appears in the conversation as what the customer chose, and a form is never dumped as raw data.
    • One turn at a time, in order — messages arriving in quick succession are answered in the order they were sent, once each, and a platform re-delivering the same message no longer produces a second reply.
    • Reactions are not messages — an emoji reaction no longer triggers a reply.
    • The assistant collects what it needs itself — when a detail is required before your team can be brought in, the assistant asks for it rather than blaming your agents for being unavailable.
    • Your customer's name is filled from their platform profile, so your team sees who they are talking to.
    • A reply already paid for is never discarded — transient platform failures are retried rather than thrown away.
    • Usage figures repair themselves — gaps between usage records are detected and repaired, so your chart and your bill agree.
    • Rejections come back as rejections — an invalid request returns a clear client error with the field named, rather than a generic failure.

Gydr 2

20 releasesJune 27 – August 14, 2026

  1. v2.8.1

    Your assistant can now ask visitors to rate something right inside the chat, and the REST API importer accepts far more API file formats than before — including older ones — without timing out on large documents. The Visitors list also gets more honest: it ranks people by when they last actually said something, and no longer fills up with people who opened the chat and never typed.

    3 new7 changed5 fixed

    What's New3

    • Your Assistant Can Ask for a Rating

      When a graded opinion is what's needed — how satisfied someone was, how good a recommendation felt — your assistant can now ask for it directly in the chat, as either a five-star scale or a simple thumbs up/down. It picks the right one for the question and shows the answer back in the summary.

    • Imports Accept Older and Newer API Files

      The REST API importer now reads Swagger 1.0 and 2.0 alongside OpenAPI 3.0 through 3.2. If your API documentation predates the modern format, you no longer have to convert it first.

    • Large API Files No Longer Time Out

      Importing a big API document used to fail partway through if it took too long. Generation now runs in the background while the page keeps you posted, so large documents finish instead of giving up. Your tool's name and description are also drafted for you from the file, alongside the endpoints.

    Changed7

    • Visitors are ranked by their last message, not their last visit — someone who just wrote to you now sits above someone who merely reopened the chat window. The column is now labelled "Last Activity."
    • Only real conversations appear in Visitors — opening the chat widget without typing anything no longer creates an entry, so the list reflects people you actually talked to.
    • The "Referrer" field is gone from visitor details — it could not be captured reliably and showed a value that misdescribed where the visitor came from. Nothing replaces it.
    • In-chat forms use the whole width available — questions sit beside their answers, prompts line up with the choices next to them, and the layout no longer changes based on how many options a question happens to have.
    • Enter moves you through a form — visitors can complete each step from the keyboard. Confirmations that end a conversation deliberately don't respond to Enter, and put the emphasis on going back rather than on ending.
    • Forms show a receipt after submitting — instead of a single line of acknowledgement, the visitor sees what they sent.
    • Handing over to a person explains itself — the confirmation now tells the visitor their recent messages travel with them, and the transfer options show the name you gave each team rather than an internal label.

    Fixed5

    • Half-star ratings can be submitted — picking 3.5 stars is now accepted rather than rejected.
    • The browser tab icon follows the widget you're viewing — switching between widgets no longer leaves the previous one's logo on the tab, and a widget without a logo no longer inherits another's.
    • Typing in Chinese, Japanese or Korean no longer sends half a word — pressing Enter to confirm a character no longer sends the message mid-composition, in both the chat box and in forms.
    • Long words in a form question stay readable — an order reference or a long link in a question is no longer cut off past the edge of the panel.
    • A question isn't shown twice — a single-question form no longer repeats the same sentence as both its heading and its question.
  2. v2.8.0

    This release hands you control over the disclaimers your assistant writes on its own, a far faster way to set up REST API tools, and a more disciplined assistant overall — one that stops showing products it didn't recommend and stops re-asking questions you've already answered.

    3 new3 changed7 fixed

    What's New3

    • Turn Off Disclaimers Your Assistant Adds On Its Own

      Every AI model has a habit of appending advice like "consult a professional," even when you never wrote it. If you operate in a regulated space — medical, financial, or legal — you can now switch that off per area from your widget settings. Your own disclaimers from your knowledge base are never touched; only the ones the model invents. It's opt-in, so leaving it alone changes nothing.

    • Set Up REST API Tools by Importing Your OpenAPI File

      Instead of typing every endpoint by hand, import your OpenAPI file and watch your endpoints fill in, with names and descriptions drafted for you. Review, adjust, install. What took many minutes now takes a fraction of it.

    • Choose the Model That Drafts Your Tool Configuration

      If you bring your own model, you can now pick which one drafts your REST API tool configuration. If it can't be reached, you get a clear error instead of a silent swap to a different model.

    Changed3

    • Dynamic forms use the full width of the chat — questions, choices, sliders and ratings now lay out for the space they actually have, instead of being squeezed into a narrow column.
    • Faster replies — answers arrive quicker, most noticeably on questions that draw heavily on your knowledge base.
    • More accessible settings switches — switch rows are now a full tap target and read their description to screen readers.

    Fixed7

    • The product carousel no longer contradicts the reply — when your assistant says it can't find a match, it no longer shows a row of unrelated products underneath. What appears is what the reply actually recommends.
    • Answers stay honest when the match is weak — rather than presenting a loosely related item as though it answered the question, your assistant now says it doesn't have that information.
    • Disclaimer settings survive a save — the switches keep their state when you save and publish.
    • Forms no longer re-ask what you've already answered — including when the follow-up is worded differently.
    • Form answers now pull from your knowledge base — answering a form question that needs information from your content now retrieves it.
    • Replies in non-Latin scripts are no longer dropped — answers in Chinese, Malay and other non-Latin scripts are delivered correctly.
    • OpenAPI import no longer leaves a blank first endpoint — the seeded row now fills in properly.
  3. v2.7.2

    This release makes your chatbot's product recommendations more complete and more consistent — what the reply lists and what the recommendation panel shows now always match. Widget domain restrictions are enforced properly again, White Label emails carry your own branding, and your custom domain logos now live where you manage the domains themselves.

    2 new3 changed3 fixed

    What's New2

    • Your Branding in Every Email

      Emails sent to your team and your customers now use your White Label brand colours and logo, so the whole experience stays yours from the chat window through to the inbox.

    • Domain Logos Where Your Domains Live

      The logo shown on your custom domain is now managed directly in Custom Domains, alongside the domain it belongs to. Uploading and replacing a logo is more reliable.

    Changed3

    • Richer answers — your chatbot now shows as many relevant results as the question genuinely calls for, instead of settling on a fixed few. Ask about a category with many matches and you'll see more of them.
    • Up to 10 product recommendations — product answers now surface up to ten items, and the recommendation panel always shows exactly the products the reply talks about — no extras, none missing.
    • Larger REST API connections — a REST API connection now supports up to 100 endpoints and a bigger configuration, so you can bring more of your service into one connection.

    Fixed3

    • Widget domain restrictions apply correctly — if you've restricted your widget to specific domains, that restriction is now enforced against the page the widget is embedded on. Clearing the list restores unrestricted access as expected.
    • Recommendation panel matches the reply — the panel no longer shows products the reply never mentioned, or leaves out one it did.
    • Accurate API responses — when a request is rejected, the API now returns the matching error status instead of a success response carrying an error inside it.
  4. v2.7.0

    This release gives White Label customers full control of their branding on the screens and emails their own users see, makes the "talk to a human" button trustworthy — it now appears only when someone is genuinely there to answer — and hardens website imports so one problem page can no longer cost you an entire crawl. It also closes several issues affecting the Conversation API and chatbots running on your own model keys.

    1 new6 changed8 fixed

    What's New1

    • Your Own Logo on Sign-In Screens and Emails

      If you use a custom domain, you can now upload a dedicated logo for your sign-in, registration, and password-reset screens, and a separate one for the header of the emails your team receives. Each is previewed at the shape and on the background it will actually appear on, so you can see how it lands before you save. Where no logo is uploaded, your account name is used — your users never see ours.

    Changed6

    • A refreshed Gydr logo — new brand mark across the website, console, and emails.
    • Handoff is offered only when someone can answer — If your help desk integration reports no agents online, the chatbot no longer shows a connect button. Instead it shares the alternative contacts you configured, so your visitors are never handed a button that leads nowhere.
    • Clearer import results — Items that could not be saved during a website import are now reported back to you instead of silently disappearing from the count.
    • Failed imports keep their work — When part of an import cannot be recovered, the extracted results are retained rather than cleared straight away.
    • Consistent Integration API usage recording — Read requests to the Integration API now record credit usage in line with the documented per-request billing. Previously some read requests reported usage that was never actually recorded.
    • More useful API error responses — Conversation API failures now come back with a specific status and message naming what went wrong, instead of a generic error.

    Fixed8

    • Public API requests were rejected — API keys were being turned away with an authorization error on the Knowledge and Conversation endpoints. Fixed.
    • A single bad page could destroy an entire website import — One document that failed validation would take the whole crawl down with it. Imports now save everything they successfully extracted.
    • Model reasoning could appear in chatbot replies — When running on your own model keys, some models' internal reasoning text could reach visitors. It is now removed in every form we have seen.
    • Rating a conversation could fail with an unexplained error — Rating a conversation that had not yet ended now returns a clear message, and is never charged for.
    • Switching accounts could get stuck — An unsaved-changes prompt could trap you on the previous account's page. Account switching now completes cleanly.
    • Emails showed a broken logo image — The header image in Gydr-branded emails failed to load. Fixed.
    • Test page buttons behaved unexpectedly — Clicking a connect-to-human or end-session button on the widget test page no longer sends a stray message or spends credit; it explains what would happen for a real visitor instead.
    • Knowledge Base icon alignment — The icon sat off-centre on the collapsed console sidebar.
  5. v2.6.2

    This release clears up the custom domain setup experience for White Label accounts. When a certificate can't be issued for your domain, you now get the real reason and the exact record to add — and removing a domain to try again actually works. We've also fixed two ways your chatbot could give a less accurate answer than your content deserved.

    4 fixed

    Fixed4

    • Custom domain certificate errors now name the real cause — When a certificate couldn't be issued for your domain, the setup screen blamed your DNS record even when it was correct. It now identifies the actual reason and tells you exactly which record to add to fix it.
    • Removing a custom domain — Removing a domain could fail with a generic "unexpected error" and leave it in place — the very step you're asked to take before trying again. Removal now completes, and on the rare occasion something does go wrong, the message explains what.
    • Category filters no longer hide FAQ answers — Filtering your FAQs by category could exclude valid answers instead of narrowing to them. Filters now focus your results without dropping matching answers.
    • Clearer answers when your content is close but not exact — When your knowledge base held something related but not the specific detail a visitor asked for, your chatbot could imply the related item was the answer. It now says plainly when the exact detail isn't there.
  6. v2.6.0

    This release puts your brand fully in front: White Label customers can now serve the Console and their chatbot from their own domains, with your company's identity on every screen and every email. You can also raise support requests without leaving the Console, connect your chatbot to your own tools through remote MCP servers, and give teammates read-only access with the new Viewer role. The Knowledge Base has been rebuilt around how you actually work with it day to day.

    6 new6 changed5 fixed

    What's New6

    • Your Own Domains for the Console and Chatbot

      White Label accounts can now run the Console and the hosted chatbot on their own subdomains — your customers and your team never see a Gydr address. A guided four-step setup walks you through proving ownership and pointing your DNS, with live checks at each step so you can see exactly where you are. Invitation, password-reset, and verification emails all carry your branding, and your sign-in screens inherit your logo and colors.

    • Support Requests from the Console

      You can now open a support ticket directly in the Console, attach files, and follow the whole conversation in one thread. You'll get an email when we reply, and tickets you've stopped responding to close themselves automatically so your list stays meaningful.

    • Connect Your Chatbot to Remote MCP Servers

      The new MCP Connector lets your chatbot use tools you already run. Point it at a server URL and it works out how to authenticate on its own — no keys to copy, no configuration to fill in.

    • A Read-Only Viewer Role

      Invite people who need to see what's happening without being able to change it. Viewers can read your analytics, conversations, visitors, widgets, skills, and knowledge, while billing, team management, and account settings stay hidden entirely.

    • Exclusion Rules and Multi-Value Custom Attributes

      Custom attributes on products and locations now support multiple values per field, and you can mark a set of values as an exclusion — items carrying them are withheld from matching visitors instead of being recommended.

    • Bring Your Own Model: Requesty

      Requesty joins the list of providers you can connect with your own API key.

    Changed6

    • The Knowledge Base, rebuilt — Knowledge is now split into Library and Records so you can tell your knowledge bases apart from what's inside them at a glance. Content editors open in a wide side panel instead of a cramped modal, with field markers showing which fields your chatbot actually reads.
    • Form answers now appear in the transcript — When a visitor answers an in-chat form, their answer shows as a proper message bubble in the conversation, in live chat and in history alike.
    • Clearer team roles — The Team settings page now explains what each role can do from a single source, so what you read is always what's enforced.
    • Filter visitors by widget — The Visitors inbox can be narrowed to a single widget.
    • Rewritten documentation — The documentation site has been rebuilt end to end: corrected facts, cleaner navigation, working in-page links, and credit costs collected in one place.
    • Every mutating button now shows progress — Buttons that save, delete, or publish show a real in-flight state so you're never left wondering whether a click registered.

    Fixed5

    • Widgets that wouldn't open — Widgets whose skills referenced a plugin could fail to load with a "Widget not found" error. They now open normally.
    • The mobile chat panel — On phones, opening the chat could scroll the page underneath, expose the host page behind the panel, or leave the panel misaligned after the keyboard appeared. The panel now stays put and the page behind it stays locked.
    • More reliable handoffs to a human — Requests to speak with a person now consistently produce a handoff, and stale handoffs no longer linger on a conversation.
    • More accurate answers when information is missing — Your chatbot is better at saying it doesn't know rather than filling the gap, and it no longer discards a correct answer that was already grounded in your content.
    • Console focus and scrolling — Menus now move focus to the right place, the space bar no longer scrolls the page from an image uploader, and navigating between pages no longer leaves the page unable to scroll.
  7. v2.5.5

    This release makes the Integration API a true two-way sync channel for your knowledge bases — you can now upsert, reconcile, and re-embed your catalog using your own product keys. The Widget Studio canvas also gets a clearer picture of how messages actually reach your widget, with per-skill resource lanes and a safer editing boundary.

    2 new3 changed2 fixed

    What's New2

    • Sync Your Catalog Through the Integration API

      Products, FAQs, and locations can now be upserted and reconciled through the Integration API using your own identifiers (SKU or external ID) — so your PIM or backend can push updates to rows it already owns instead of only creating new ones. Partial updates are safe by design: fields you don't send are left untouched, and custom attributes merge per key. You can also trigger a knowledge base sync and poll its status directly with your API key, making newly written content answerable in chat without opening the Console.

    • Channel View in Widget Studio

      The Studio canvas now shows the three ways a conversation reaches your widget — Website Embed, Hosted Page, and the Integration API — each with its own path in and out. Disabling your hosted page is reflected live on the canvas.

    Changed3

    • Per-skill resource lanes in Widget Studio — Each attached skill now gets its own lane holding exactly its knowledge bases and plugins, instead of one shared column of your whole catalog. Shared resources are shown in every lane that uses them, and lanes fold to keep large setups readable.
    • Safer editing in Widget Studio — The canvas now edits only the widget you're looking at. Account-wide resources (skills' knowledge wiring, plugins, and more) are shown read-only, with each panel linking to the page that owns it — so a change in one widget's Studio can no longer affect your other widgets.
    • Clearer message flow on the canvas — The conversation loop now enters and exits the core at fixed points, with a tidier layout for the visitor, channel, and response paths.

    Fixed2

    • Knowledge writes through the Integration API — Creating or updating knowledge items with an API key could fail with a server error. Writes now complete reliably, and validation errors name the exact field that needs fixing.
    • Widget Studio layout drift — Stale hand-dragged positions are discarded when the canvas layout is updated, so the diagram always renders its intended shape.
  8. v2.5.4

    This hotfix makes the embedded chat widget more dependable on your website. Visitor contact details now arrive exactly as entered, the widget rides out temporary network hiccups instead of failing silently, and visitors always see a clear status instead of an endless loading spinner.

    1 changed5 fixed

    Changed1

    • Hardened widget security — Strengthened the protections around the embedded widget's handling of visitor information and custom branding assets, backed by an expanded automated test suite that guards every supported embedding method.

    Fixed5

    • Visitor details preserved exactly — Visitor names, emails, phone numbers, and company names containing spaces or hyphens were being altered before reaching your conversations (for example, a two-word name arriving joined together). Details you pass to the widget now come through exactly as entered.
    • Widget recovers from network hiccups — A brief connection blip while your page loaded could previously leave the chat widget missing until the visitor refreshed. The widget now retries automatically and loads normally.
    • No more endless loading spinner — If the chat genuinely can't load (for example, a restrictive site configuration), visitors now see a clear "Chat is currently unavailable" message instead of a spinner that never resolves.
    • Chat loads in page-builder previews — The widget could get stuck loading when your page ran inside certain preview or sandboxed environments (such as site-builder preview panes). It now starts correctly there too.
    • Custom launcher icons fall back gracefully — An invalid custom icon or an image that fails to load now shows the default chat icon instead of a broken graphic.
  9. v2.5.3

    This patch restores chatbot embedding for websites using the standard script snippet, and quiets an unnecessary browser prompt inside the Console on some mobile devices. If your embedded chatbot was not appearing, this update brings it back automatically — no change to your embed code is needed.

    2 fixed

    Fixed2

    • Embedded chatbot not appearing — Websites embedding the chatbot with the standard script snippet (the tag with your API key) could fail to show the chat bubble or chat window. The snippet now activates reliably again. Integrations that start the chatbot from their own code were not affected.
    • Unnecessary download prompt in the Console on mobile — On some mobile browsers, opening the Console could trigger a small, harmless file-download prompt that overlapped the support chat bubble. Payment-security scripts now load only on billing pages, removing the prompt everywhere else while keeping payment fraud protection fully intact.
  10. v2.5.2

    This patch restores the embedded chat widget for visitors on older Safari and iOS versions, alongside two smaller polish fixes. We recommend all customers on v2.5.1 take this update — no action is needed on your side; embedded widgets pick up the fix automatically.

    3 fixed

    Fixed3

    • Embedded widget on older Safari and iOS — On Safari and iOS versions below 16.4, the embedded chat widget could fail to load and appear blank after the v2.5.1 update. The widget now loads correctly on these browsers.
    • More accurate replies to combined requests — When a visitor's message combines a question with an action in the same turn, the assistant now verifies its answer more reliably instead of over-correcting it.
    • Account switching visuals — When switching between accounts in the Console, the loading overlay now shows the look of the account you're switching to, not the one you're leaving.
  11. v2.5.1

    A small reliability release for the chat experience. Forms your chatbot presents now behave the way a human would — ask, then wait — and replies render more cleanly across languages and regions.

    3 fixed

    Fixed3

    • Forms no longer arrive with the answer attached — When your chatbot asks a question with a form (a clarifying choice, a location confirmation, a booking detail), it now waits for the visitor's response instead of sometimes answering in the same message. Asking and answering are now cleanly separated turns.
    • Product prices follow the visitor's locale — Prices shown in chat now format according to each visitor's own regional settings.
    • Cleaner replies — Removed rare cases where internal notation could appear inside a chat reply.
  12. v2.5.0

    This release gives you two big new ways to shape your chatbot. Define your own custom attributes on your product and location knowledge bases — like size, material, dietary tags, or amenities — and let your chatbot filter and answer by them. And meet Widget Studio, a visual canvas where you can see and wire your entire widget — skills, knowledge, and integrations — as an interactive map. We've also refreshed knowledge editing and sharpened your chatbot's answers.

    2 new1 changed3 fixed

    What's New2

    • Custom Attributes for your knowledge bases

      You can now define your own attributes on product and location knowledge bases — anything that matters to your catalog, like size, color, material, dietary tags, or store amenities. Once defined, your chatbot can filter and answer questions using them, so a visitor asking for items with a specific feature gets exactly the right matches. Author attributes directly in the console, bring them in through CSV import or the integration API, or have them picked up automatically when you crawl your website.

    • Widget Studio — a visual widget editor

      Build and understand your widget on an interactive canvas. Widget Studio lays out your widget, its skills, knowledge bases, and integrations as connected nodes you can see and wire together at a glance, then save and publish when you're ready. It's a clearer, more visual way to configure everything your chatbot uses to answer.

    Changed1

    • Redesigned knowledge editing — The forms for adding and editing products, FAQs, documents, and locations are now multi-column workbenches, making it faster and easier to fill in and review your knowledge base content.

    Fixed3

    • Global outlet questions answer correctly — When a visitor asks a general question about all your locations (like which outlet has a particular feature), your chatbot no longer narrows the answer to a single saved location.
    • Deleted knowledge is fully removed — Items you delete from your knowledge base, including in bulk, are now cleared from search right away, so removed content no longer appears in answers.
    • More accurate, better-grounded answers — A round of answer-quality improvements: your chatbot stays closer to your knowledge base, matches product requests more precisely, and won't add unsolicited disclaimers.
  13. v2.4.1

    This update reorganizes your billing experience for easier navigation, improves credit efficiency, lets you switch back from a custom plan to a standard tier, and fixes several display and billing accuracy issues.

    2 new2 changed3 fixed

    What's New2

    • Restructured Billing Settings

      Billing is now organized under a dedicated Settings section with separate pages for managing your subscription plan and purchasing credit packs. Partner pricing now shows slashed comparison pricing so you can see your savings at a glance.

    • Switch from Custom to Standard Plan

      If you're on a custom plan, a new Restore button lets you switch back to a standard tier in one click.

    Changed2

    • Improved Credit Efficiency — Your credits now go further, with more chat messages and AI-powered responses delivered per credit.
    • Clearer Payment Summaries — Plan confirmation and upgrade modals now show a more detailed breakdown of what you'll be charged and when.

    Fixed3

    • Partner Pricing Display — Partner discount pricing now shows correctly only on the primary account, with centered discount cards and banners, and updated discount text across all supported languages.
    • Custom Plan Banner — Custom plans now appear as a hero banner above the standard tier options, and the "current plan" badge updates correctly after switching plans.
    • Mid-Cycle Plan Changes — Proration during mid-cycle plan changes now uses invoice-level adjustments for accurate billing.
  14. v2.4.0

    This release brings a complete redesign of how your chatbot collects visitor information. Forms are now step-by-step with a review screen, date and time questions get themed pickers, and your chatbot can offer only the dates and times your services actually have available. We've also added custom subscription plans, moved BYOM to Growth+, and shipped important fixes for chat accuracy and product editing.

    2 new4 fixed1 breaking

    Breaking Changes

    These may need a change on your side.

    • BYOM is now Growth+ — Bring-your-own-model is available on Growth+ plans and above. If your account is on a Starter plan or below using BYOM, you will need to upgrade to continue using this feature. BYOM access stops automatically on downgrade, and your chatbot will continue answering using Gydr's built-in model.

    What's New2

    • Dynamic Forms — A Complete Redesign

      Forms in your chatbot have been rebuilt from the ground up. Instead of a wall of questions, your visitors now move through one question at a time with a progress bar, then review all their answers on a confirmation screen before submitting. Date and time questions come with themed calendar and time-slot pickers, making appointments and bookings feel natural. Your chatbot draws available dates and times directly from your connected tools and services — visitors only see slots that actually exist. Answers are displayed in both the visitor's local time and UTC, so dates and times are always unambiguous. Submitted forms now appear as clean, structured cards in your conversation transcripts.

    • Custom Subscription Plans

      You can now configure fully custom subscription plans with bespoke credit allocations, seat limits, and feature sets — perfect for accounts that don't fit the standard tier model. Billing for custom plans is production-ready from the start, so there are no unpaid activation gaps.

    Fixed4

    • More accurate chat responses — Your chatbot now correctly declines with "I don't have that information" when it genuinely cannot find an answer, instead of guessing or making something up. It no longer claims your knowledge base is missing information when it actually has it. And it stays focused on your content — it won't volunteer unsolicited professional-disclaimer boilerplate unless your own knowledge base already includes that language.
    • Product editing: clearer errors and unblocked legacy items — When saving a product fails, you now see exactly which field needs attention and what the character limit is, rather than a generic error. Products written before character limits were enforced can now be edited again without being blocked by their own legacy content.
    • Plugin configurations restored — Some plugin configurations that were set up before early July appeared to vanish silently. They have been recovered and are now visible again in your plugin settings.
    • Form behavior improvements — When your chatbot creates more questions than fits in one form, the overflow now carries over to the next form instead of being lost. Earlier answers no longer interfere with the chatbot authoring the next form's questions.
  15. v2.3.0

    Your chatbot now answers in the language your visitor writes in, every time — even when your knowledge base is written in a different language. Goodbyes are recognised in any language, so the end-of-chat prompt appears when it should. We've also fixed an editing surprise where renaming a chatbot in a saved configuration could rename your live one, and custom-plan customers can now see their negotiated details right in the dashboard.

    1 new1 changed6 fixed

    What's New1

    • Your negotiated plan details, in the dashboard

      If you're on a custom plan, Settings → Billing now lists the items you agreed with us, alongside your usual plan limits. Available in English, Malay and Chinese.

    Changed1

    • Goodbyes recognised in any language — when a visitor says they're done, the end-of-chat confirmation now appears regardless of the language they're chatting in.

    Fixed6

    • Replies match your visitor's language — your chatbot no longer drifts into the language your knowledge base happens to be written in. A visitor who writes in Malay gets a reply in Malay, even when your content is in English.
    • No more stray conversation labels — replies occasionally began with leftover formatting like a speaker label. That's gone, in every language.
    • Renaming a chatbot stays in your draft — editing a saved configuration's name or description no longer changes your live chatbot. Your published chatbot keeps its name until you publish the change.
    • Access restored correctly after a payment issue — custom-plan accounts now regain their full credits and features as soon as payment is resolved.
    • Partner-managed accounts stay active — accounts remain usable after a partner is removed from them.
    • Cancelling a custom plan works on every base plan — cancellation no longer gets stuck for some accounts.
  16. v2.2.0

    This release brings flexible custom billing arrangements for partners and enterprise accounts, plus meaningful improvements to how accurately your chatbot answers questions from your knowledge base. We've also fixed a multilingual greeting issue so visitors are routed correctly no matter what language they open with.

    1 new1 changed2 fixed

    What's New1

    • Custom monthly commitment billing for partners & enterprise

      Partners and enterprise accounts can now be set up with tailored monthly commitment plans that match a negotiated arrangement, with reliable activation and re-pricing. If your business has a custom agreement, your billing now reflects it cleanly.

    Changed1

    • More accurate answers across your whole knowledge base — your chatbot now applies a consistent relevance standard to every knowledge type (products, FAQs, documents, and locations), so visitors get stronger, on-topic answers and fewer loosely-related results slip through.

    Fixed2

    • Multilingual greetings routed correctly — fixed an issue where an opening greeting in some languages could be misrouted, so conversations now start smoothly regardless of the visitor's language.
    • Steadier answer generation — resolved a rare case that could interrupt how the chatbot assembled a response, making replies more consistently reliable.
  17. v2.1.0

    This release gives you full control over how your chatbot changes go live. Build and refine your widget as a draft, save named versions you can return to, lock the ones you want to protect, and publish exactly when you're ready. We've also sped up chatbot responses and fixed several reliability issues across chat and the test tools.

    2 new2 changed4 fixed

    What's New2

    • Save and publish workflow for your chatbot

      You can now edit your chatbot as a draft without affecting what your visitors see. Save your work as named versions, revisit or restore any earlier one, and publish the exact version you want live — all from a redesigned editor that clearly separates what's LIVE from what you're EDITING.

    • Named, restorable versions with locking

      Give each saved version a meaningful name so you always know what it contains. Lock a version to protect it from accidental changes or deletion, and rename published versions to keep your history tidy.

    Changed2

    • Faster chatbot replies — conversations feel noticeably quicker, especially during back-and-forth exchanges.
    • Redesigned widget editor — a clearer two-row layout separates your live chatbot from your in-progress edits, so you always know exactly what your visitors are seeing.

    Fixed4

    • Test widget no longer gets stuck — the preview no longer hangs on "Thinking…" when a test message fails, so you can keep iterating.
    • More reliable in-chat forms — forms your chatbot presents mid-conversation now recover cleanly, so visitors aren't left on a stalled step.
    • Better answers across languages — your chatbot more reliably draws on your knowledge base when a visitor writes in a different language than your content.
    • Steadier connections — the chatbot recovers gracefully from brief connectivity blips during a conversation.
  18. v2.0.2

    This release makes testing your chatbot before launch much more powerful. You get a redesigned standalone test page, a new debug view that shows how your chatbot reached each answer, the ability to export test transcripts, and test runs that use your real integrations. We've also sped up list and analytics loading and improved multilingual conversations.

    3 new3 changed3 fixed

    What's New3

    • Redesigned Widget Test page

      A dedicated, full-page space to chat with your widget exactly as your visitors will — so you can validate its answers, personality, and knowledge before you go live.

    • Debug view for test conversations

      A new panel breaks down how your chatbot handled each test message — the knowledge and tools it drew on, step by step — so you can pinpoint what to refine in your skills and knowledge base.

    • Export test conversations

      Download a full transcript of any test chat to review, share with your team, or keep for reference.

    Changed3

    • Test with your real integrations active — test conversations now run your connected plugins and integrations, so what you see matches what your visitors will experience.
    • Faster lists and analytics — dashboard list and analytics views load noticeably quicker.
    • Polished test surface — a cleaner, more usable layout across the widget test page, including a scrollable debug panel for longer conversations.

    Fixed3

    • Personality applied more reliably — your chatbot now follows its configured personality more consistently.
    • More relevant product suggestions — the chatbot is more selective about when it surfaces products, so suggestions stay on-topic.
    • Better multilingual handling — improved language detection for conversations that mix Chinese, Japanese, or Korean characters with Latin text.
  19. v2.0.1

    This release tightens how your chatbot draws on your knowledge base and gives you more visibility while configuring its skills. Answers stay more faithful to your content — especially across languages — and a new context meter shows you exactly how much of your chatbot's instruction space you're using as you edit.

    1 new1 changed3 fixed

    What's New1

    • See your instruction space as you build

      When configuring your chatbot's skills, you can now see a live meter of how much of its available instruction space each skill, plugin, and setting takes up — updating as you type. It's easier to tell at a glance when you're adding a lot, so you can keep your chatbot focused.

    Changed1

    • Skill instructions followed as written — The instructions you give each skill are now applied exactly as you wrote them, so your chatbot behaves more predictably and stays on-message.

    Fixed3

    • More reliable knowledge-base answers — Your chatbot now consistently draws on your FAQs, documents, and other knowledge whenever it's helping a customer, so fewer relevant answers slip through.
    • Exact details across languages — When your chatbot replies in a different language than your knowledge base, technical details like prices, codes, and specifications now stay precise and unchanged.
    • Product links preserved — Links to your products now reliably appear in your chatbot's answers.
  20. v2.0.0

    Major release

    Our biggest release yet, led by two headline features. Dynamic Mode lets you choose a set of AI models and have your assistant automatically pick the best one for every message. Plugins connect your chatbot to the external tools and data you already use — including a new commerce connector for your online store. On top of that: a memory for returning visitors, finer control over your knowledge base, clearer cost visibility, and an assistant rebuilt to answer more accurately and stay on topic.

    7 new11 changed8 fixed

    What's New7

    • Dynamic Mode — the right AI model for every message

      Choose a set of AI models for your widget and turn on Dynamic Mode. Your assistant routes each message to the best-fitting model automatically — keeping everyday questions fast and economical while sending harder ones to a stronger model. If a model can't answer a question well, the conversation automatically escalates to a more capable model from your set, and if your knowledge base is in another language, Dynamic Mode picks a model that can read it. You'll see a clear cost indicator on every model so you can choose with confidence, and Dynamic Mode conversations are billed at a discount.

    • Plugins — connect your chatbot to the tools you already use

      Plugins let your chatbot reach beyond its knowledge base. Connect your own REST API (with flexible authentication, custom headers, and multiple endpoints), or sync answers from a Google Sheet — now with per-connection instructions and column mapping so the assistant understands your data. A redesigned catalog with a guided, tabbed setup makes connecting straightforward.

    • Sell from your store with the Shopify connector

      A new Shopify Storefront plugin lets your chatbot help shoppers search your products, check availability, build a cart, and hand off to your store's secure checkout — plus look up their own past orders when signed in. Browsing is on by default; cart actions are off until you enable them, and the assistant never places or pays for an order: checkout always stays on Shopify.

    • Visitor memory

      Your chatbot can now remember helpful context about returning visitors across conversations, so repeat customers don't have to start over. It's off by default and controlled per widget with a clear consent toggle, and you can review or erase what's remembered at any time from the visitor's profile.

    • Finer control over your knowledge base

      Turn individual documents on or off without re-uploading — and without triggering a needless re-sync. Edit your document content chunk by chunk directly in the console, with helpful best-practice tips and a clear character limit, for precise control over exactly what your assistant uses.

    • See what each conversation costs

      Per-message and per-conversation credit costs now appear in the console, so you can understand exactly where your usage goes.

    • Try it out, on any plan

      The "Try it out" sandbox is now open on every plan, including Free, so you can test your live chatbot before you publish it.

    Changed11

    • "Plugins" is the new home for your integrations — what used to be called MCP Tools is now Plugins, with redesigned, easier-to-navigate setup pages and brand logos for each connector. Your existing configuration carries over.
    • More features on Starter — Bring Your Own Model and Plugins are now available starting on the Starter plan, and human-handoff integrations are now a Starter-and-above feature.
    • More accurate, on-topic answers — the assistant now answers strictly from your knowledge base rather than general training, and declines honestly when the answer genuinely isn't there instead of guessing.
    • Answers in your visitor's language — replies follow the visitor's language throughout a conversation, including switches mid-chat and questions asked in a different language than your content.
    • Recognizes your own services — if your business sells technical or specialized services, the assistant now treats questions about them as in-scope and answers from your knowledge base instead of refusing.
    • Cleaner catalog browsing — a broad "what do you have?" now lists your products cleanly instead of an awkward "no matches found" hedge.
    • A cleaner chat experience — a refined thinking indicator while the assistant works, and a message composer that adapts between touch and desktop.
    • Smarter website crawling — the crawler now renders modern, JavaScript-heavy pages before importing and keeps an item even when one field runs long, so more of your site's content makes it into your knowledge base.
    • Personality builder polish — the Make-your-own personality flow is clearer, previews each style as a sample conversation, and validates before you save.
    • Knowledge-base table refinements — column-aligned rows, per-type color accents, and save confirmations.
    • Stronger compliance audit logging — broader, more consistent security audit records across account and billing actions.

    Fixed8

    • Paid plans no longer get stuck on Free — fixed a payment-processing race that could leave a paid subscription showing as Free after an out-of-app or interrupted payment; affected accounts now self-heal automatically.
    • Payments needing extra verification now complete — repaired checkout for cards that require an additional verification step from the bank.
    • Custom personalities apply reliably — a saved custom personality (Growth and above) now correctly shapes your live chatbot's replies.
    • Accurate billing allocations — fixed monthly credit-allocation drift for partner-managed accounts.
    • Forms behave reliably — location requests, clarifying questions, and end-of-chat prompts now appear and submit dependably, never loop, and always offer a way out.
    • Location is remembered — a visitor whose location is already known is no longer re-asked on every nearby-search.
    • Messages arrive complete — the chatbox now waits for the full response, so the "thinking" indicator never gets stuck.
    • Your content survives to the answer — longer product, FAQ, and document fields are no longer trimmed before the assistant uses them.

Gydr 1

9 releasesMay 13 – June 9, 2026

  1. v1.32.1

    This release improves the accuracy and consistency of your chatbot's everyday conversations — the product guidance you write is followed more reliably, replies stay in your visitor's language, and store-location answers are more dependable.

    4 fixed

    Fixed4

    • Your product AI Context Notes are honored — The AI Context Notes you add to products (usage tips, consumption guidance, caveats) are now consistently reflected in your chatbot's answers, so it follows your guidance instead of overriding it with generic information.
    • Replies stay in your visitor's language — The chatbot no longer switches to the wrong language partway through a conversation; replies now consistently match the language your visitor is using.
    • More reliable nearby-store results — When a visitor shares their location to find the closest store, the chatbot now dependably returns nearby stores instead of asking for their location again.
    • Accurate store availability — The chatbot no longer incorrectly says a business has no physical stores (or is online-only) when store locations are available; it now surfaces the real locations.
  2. v1.32.0

    This release sharpens a few everyday touchpoints across your console and chat widget. You'll find clearer signposting for older conversations, a tidier widgets list with quick actions, and a calmer end-of-chat experience for your visitors.

    1 new3 changed

    What's New1

    • Know when older conversations have aged out

      Your Visitors area now clearly explains when a conversation has passed its retention window, so it's obvious why an older transcript is no longer available — no more guessing. Retention length depends on your plan.

    Changed3

    • Tidier widgets list with quick actions — Each widget now shows its agent avatar at a glance, and a new per-row menu puts common actions a single click away.
    • Calmer end-of-chat for visitors — When a conversation times out, your chat widget now shows a single, clear "conversation ended" state right where your visitor was typing, with a one-tap way to start a fresh conversation.
    • Interface polish — A round of accessibility, motion, and visual refinements across the console.
  3. v1.31.4

    This release makes your chatbot more resilient — it stays focused on your business and shrugs off attempts to push it off-topic or manipulate it. Signing up and signing in is smoother, with the system automatically picking up where you left off if a sign-up gets interrupted. We've also refreshed the Audit Log and fixed a couple of rough edges in links and filtering.

    1 new2 changed2 fixed

    What's New1

    • Stronger chatbot safeguards

      Your chatbot now does an even better job staying on the topics that matter to your business. It more reliably keeps conversations focused on your products, services, and knowledge base, and is more resistant to attempts to steer it off-topic or manipulate its responses — so visitors get answers that stay on brand and on point.

    Changed2

    • Smoother sign-up and sign-in — Creating an account and signing in is now more reliable. If a sign-up gets interrupted partway through, the system automatically picks up where you left off the next time you return, so you're no longer left stuck mid-process.
    • Refreshed Audit Log — The Audit Log has a cleaner, easier-to-scan layout with clearer event details, making it simpler to review account activity at a glance.

    Fixed2

    • Audit Log filtering — Filtering the Audit Log by action now works correctly and returns the matching events instead of coming up empty.
    • External links in chat — Links shared in the chat window now open properly in a new tab instead of failing to load.
  4. v1.31.3

    This release sharpens the console experience — especially on mobile — and makes inviting teammates rock-solid. Slide-out panels now feel consistent everywhere, the knowledge base is easier to get around, and a handful of small layout quirks are gone.

    4 changed3 fixed

    Changed4

    • Consistent slide-out panels — every slide-out panel now opens at the same comfortable width, so editing your widgets, skills, and knowledge feels uniform across the console.
    • Easier knowledge base navigation — the "Add" button now sits right where you expect it: pinned to the right on desktop and full-width on mobile for an easy tap.
    • Better mobile layouts — console pages now fill your screen properly on mobile, with cleaner, more consistent spacing throughout.
    • Automatic cleanup of abandoned invitations — team invitations that are never accepted are now tidied up automatically, freeing that email address to be invited again later.

    Fixed3

    • Reliable team invitations — inviting a teammate no longer fails partway or leaves an invite stuck in limbo; invitations now go through cleanly, every time.
    • No more layout shift — opening a panel or overlay no longer nudges the navigation bar as the scrollbar appears and disappears.
    • Cleaner crawl-job cards on mobile — website crawl-job details now wrap and reflow neatly on smaller screens.
  5. v1.31.2

    This patch release sharpens chat behavior across compound, location, and language-mixed conversations, makes Website Import recover gracefully from transient failures, and restores complete translations across the console and marketing site.

    4 changed4 fixed

    Changed4

    • More natural responses to multi-topic and location questions — When a visitor asks about several things at once, or follows up a product question with "where can I find this?", the bot now handles the conversation more smoothly without unnecessary clarifying questions or incorrect off-topic refusals.
    • Better language matching — When a visitor writes primarily in Chinese, Malay, or another non-Latin script, the bot now replies in that language even if their message mixes in a few English words.
    • Faster responses on single-topic questions — Straightforward queries no longer trigger extra processing, so simple chats feel snappier and use fewer credits per turn.
    • Complete UI translations — Console settings and the homepage chat demo now display correctly in all supported languages (English, Malay, and Chinese).

    Fixed4

    • Website Import recovers from temporary failures — The import status page used to get stuck on "Import job not found" after a transient hiccup and stop checking for progress. It now self-recovers and shows clear, friendly messages instead of raw provider error text. Large website imports also run with a more conservative concurrency to avoid hitting capacity limits.
    • Human handoff replies preserved — When the bot transfers a conversation to your team, its response message is now kept intact instead of being replaced by a generic refusal.
    • "Where can I get these?" correctly prompts for a location — Follow-up proximity questions after a product result now ask the visitor to share their location instead of replying that the topic is out of scope.
    • More consistent product and FAQ answers — Resolved several edge cases where the bot repeated the same search multiple times or included referral suggestions that were not grounded in your knowledge base.
  6. v1.31.0

    This is our biggest release yet. Power your chatbot with your own AI model provider, rebrand the entire Console as your own, and give your chatbot a custom personality you can test live before going live. The Console itself has been rebuilt from the ground up — mobile-first, faster, and now available in English, Malay, and Chinese — plus an overhauled team management experience, visitor file attachments during human handoff, and proactive email alerts when your credits run low.

    9 new10 changed10 fixed2 breaking

    Breaking Changes

    These may need a change on your side.

    • Facebook Messenger output format — for customers using the Integration API with Facebook Messenger formatting, the carousel response now uses an elements array (previously cards), and each quick reply includes a content_type field, to match Facebook's current Send API. Update any code that reads the carousel response.
    • Usage credits in the API response — the Integration API now reports credit usage broken into multiple categories (chat, retrieval, orchestration, tools, and crawler) instead of a single combined value. If your integration reads the credits field from the API response, update it to the new breakdown.

    What's New9

    • Bring Your Own Model (Growth plan and above)

      Connect your own AI provider — including OpenAI, Anthropic, Gemini, DeepSeek, Qwen, Mistral, Moonshot, OpenRouter, and Groq — and choose which model powers each of your widgets. Add and validate your connections from the new Import Models page in Settings, then pick a model per widget. When a widget runs on your own model, those conversations bill to your provider account, not your Gydr credits.

    • White Label (Growth plan and above)

      Make the Console your own. Replace the Gydr logo and accent color across the sidebar and top bar with your own branding, applied for every member of your team.

    • Custom widget personality with a live sandbox

      Give your chatbot a personality that fits your brand. Choose from eight ready-made presets or describe your own custom style, then try it out in a live sandbox to see exactly how it responds before you save. Custom personalities are available on the Growth plan and above.

    • Visitor file attachments during handoff (Preview)

      When a conversation is handed to one of your human agents, visitors can now upload files directly in the chat, and those attachments flow through to your help desk and appear in the conversation transcript.

    • Credit alert emails (Preview)

      Get notified before you run out. Configure email alerts and recipients in Settings to be warned when your credits reach a threshold you choose.

    • Instagram messaging format

      Alongside Facebook Messenger and WhatsApp Business, your chatbot can now format its replies for Instagram messaging.

    • A multilingual Console

      The Console is now available in English, Malay, and Chinese. Switch languages anytime from your profile settings.

    • Detailed usage breakdown

      Your Usage History now breaks credit usage into clear categories — chat, knowledge retrieval, tools, website crawling, and file attachments — so you can see exactly where your credits go.

    • Per-knowledge-base result tuning

      Turn on smarter result ranking for any knowledge base, with frequency presets to balance quality and cost.

    Changed10

    • A completely redesigned Console — a faster, enterprise-grade interface that's now fully mobile-friendly, with refreshed navigation and consistent, scannable list views across Knowledge Base, Skills, Widgets, and Integrations.
    • A reworked Visitors inbox — visitor conversations are now a clean master-detail inbox, with uploaded attachments shown inline.
    • Easier knowledge management — click any item to edit it, sync and select items right from the list, and get clear, plain-language messages when an item fails to sync.
    • A streamlined widget editor — widget settings are now grouped into clear tabs, each saved independently so you only update what you change.
    • A smarter website crawler — improved extraction of products, FAQs, documents, and store locations from your website.
    • Configurable low-credit warnings — set the threshold for your credit-low banner in Settings.
    • Greeting screen extras — add a line of additional information with clickable links to your widget's greeting.
    • Refreshed Account and Profile settings, plus an updated privacy and cookie-consent experience aligned with the latest data-protection standards.
    • Partner-managed accounts can now configure their own credit alerts and use the personality sandbox and other inherited plan features directly.
    • Tighter knowledge isolation — each widget now draws only on the knowledge bases and skills you've assigned to it, enforced end to end.

    Fixed10

    • Cleaner account switching — switching between accounts now fully refreshes your view, so you never see data from the account you just left.
    • Product suggestions match the reply — product cards shown alongside a chat response now line up with what the chatbot actually recommended.
    • More accurate location answers — the chatbot no longer says there are no stores when your locations can answer, and it now handles outlets that don't have a type set.
    • Correct message order on reconnect — chat messages stay in the right order and no longer duplicate or shuffle when a conversation reconnects or a session ends.
    • Sharper, more relevant answers — when there's no strong match, the chatbot no longer pads its reply with loosely-related items.
    • No more accidental sends — the chat widget now blocks a message from being sent after a session has timed out, and shows immediate feedback while a file uploads so a slow connection never looks broken.
    • Steadier reconnections — if the chat reconnects, your own messages no longer appear twice.
    • Overlay polish — dialogs, slide-out panels, and their backdrops now layer and cover the screen correctly across the Console.
    • Greeting reliability — your greeting's extra info is now saved correctly and the greeting shows even when no FAQs are configured.
    • Clearer error messages — Settings flows now show errors right inside the dialog instead of failing silently.
  7. v1.30.0

    This release sharpens how your chatbot handles everyday questions — exact counts, store hours, and more relevant product suggestions — and introduces a refreshed console navigation that's faster and smoother on both desktop and mobile.

    1 new3 changed6 fixed

    What's New1

    • Count and Time Questions, Answered Directly

      Your chatbot can now answer catalog-size questions — "how many products do you have?", "how many stores are there?" — with an exact figure. It can also tell visitors the current time and whether a location is open right now, so "are you open?" gets a straight answer.

    Changed3

    • Refreshed console navigation — A new left-edge sidebar stays compact as an icon rail and expands when you need it, with cleaner styling across the console and the Gydr logo in one consistent spot.
    • More reliable chat product suggestions — Product cards shown alongside chat replies now appear consistently and are filtered to the closest matches, so visitors see fewer loosely-related items.
    • More dependable chat session endings — Ending a chat now confirms before closing, so sessions wrap up cleanly every time.

    Fixed6

    • Accurate "how many" answers — The chatbot no longer under-counts when a visitor asks about the size of your catalog or your number of stores.
    • No more premature "I couldn't find that" — When your knowledge base has the answer, the chatbot leads with it instead of opening with a denial.
    • Cleaner chat replies — Removed occasional stray formatting characters from chatbot responses.
    • Smoother console on mobile — Fixed touch scrolling, pull-to-refresh, and a scroll-lock issue that could leave the page stuck.
    • Slide-in panels open cleanly — Configuration panels now slide in smoothly, without bouncing or layering glitches.
    • Chat session ending no longer freezes — Resolved a case where ending a session could stall.
  8. v1.29.1

    Bring Gydr's chatbot to any channel. The Conversation API can now shape responses for Facebook Messenger, WhatsApp Business, or plain text — picked per API key in your Developer settings. Your chatbot also answers questions more confidently, ranks suggestions more accurately, and handles follow-up questions better. Plus a new Rerank line in your Usage History and a couple of bug fixes.

    5 new7 changed2 fixed

    What's New5

    • Channel-Aware Responses

      Pick how each of your Conversation API keys formats its replies — Markdown (default), Plain text for SMS bridges and voice agents, Facebook Messenger with cards and quick replies, or WhatsApp Business with reply buttons, list messages, and rich product items. Connect the same chatbot to multiple channels with one API key each.

    • Map Cards for Your Outlets

      When a visitor asks for the nearest outlet, Facebook Messenger users see a proper card with the outlet photo, address, and tappable buttons (View on map, Call, Website). WhatsApp users get a native pin-on-map message with real coordinates. Beautiful out of the box — no extra work in your middleware.

    • Handoff Flag on Conversation API Responses

      Every Conversation API response (sync and callback) now includes a handoff field that tells you whether the chatbot suggested a human handoff this turn. Route to your live agent flow with one boolean instead of parsing message content.

    • Smarter Answer Ranking

      When several knowledge base items look similarly relevant, your chatbot now picks the one that actually fits the question. Especially noticeable in multilingual catalogs and rich product lists. A new Rerank line in your Usage History shows you when this kicks in.

    • Handoffs That Work on Every Channel

      The "Talk to staff" action now flows cleanly through Facebook quick replies and WhatsApp buttons. When a visitor taps it on their channel, your bridge relays it back to Gydr and the handoff fires — same protocol as the in-house widget.

    Changed7

    • Better follow-up handling — Your chatbot now connects multi-step questions more reliably. Visitors who refer back to products or topics from earlier in the chat get coherent answers instead of generic ones.
    • Answers your visitors' questions instead of deflecting — Your chatbot no longer redirects visitors to seek outside help when their question is one your knowledge base can answer. Disclaimers only appear if you've explicitly written them into your knowledge base or skill prompts.
    • Smarter suggestions for higher-budget shoppers — When a visitor indicates they're shopping in a higher price range, your chatbot now leads with premium options instead of starting from the cheapest.
    • Cleaner category handling — Category names that differ only in capitalization are now treated as the same category. No more split lists or confused follow-ups when your catalog has minor inconsistencies.
    • Conversation API responses include a new format field — Every non-error response now tells you which response shape produced it. If your code validates response shapes strictly, add format to your parser.
    • Developer tab: Response Content Format dropdown — Pick how each API key formats its Conversation API responses. Available when the Conversations: Chat scope is enabled, with a new Format column on your API keys table so you can see which key delivers which shape.
    • Richer location answers across channels — Phone numbers, map links, and websites you've configured on your outlets now show up as tappable buttons on Facebook Messenger cards and as native location messages on WhatsApp.

    Fixed2

    • Widget no longer crashes on reconnect — Reopening the chat or retrying after a credit-exhausted message used to throw an error. Visitors can now refresh and resume cleanly.
    • Smoother first-message experience after idle periods — If your chatbot was unused for a long stretch, the first message back occasionally surfaced as a timeout error. That window is now significantly smaller — visitors get a response instead of a failure.
  9. v1.28.0

    Big things in this one. Your chatbot just got a whole lot smarter — it actually follows the conversation instead of just matching keywords. We rebuilt the widget so it auto-updates on every release (no more copy-pasting embed codes!). The Console finally works great on your phone, and visitors can now close their browser and come back hours later to pick up right where they left off.

    8 new4 changed6 fixed2 breaking

    Breaking Changes

    These may need a change on your side.

    • New chatbot response engine — Your chatbot now runs on a rebuilt response engine that delivers smarter, streaming replies. If you built custom integrations against the previous response engine, you may need to update them. The Conversation API surface is unchanged.
    • Custom chatbox URL override is deprecated — The chatboxUrl override on the widget embed snippet is no longer supported. We've tightened how the chatbox loads to reduce the risk of compromised embeds. Remove the override from your embed code.

    What's New8

    • Your Chatbot Just Got Better at Thinking

      Your chatbot now understands what your visitors actually want. It handles multi-step questions, knows when to say "I'm not sure," and streams responses live so visitors see the words appearing instead of a loading spinner. If someone interrupts themselves mid-question or bounces between topics, the chatbot keeps up. No more blank responses or confused replies.

    • Conversations That Survive Anything

      Accidentally close the tab? Refresh the page? Come back tomorrow? No problem. Your visitors now pick up exactly where they left off — the full history restores instantly. And from the Console, you can search and browse every past conversation, so your team never loses context.

    • Mobile Console Is Here

      Manage your chatbot from anywhere. We added a proper navigation drawer for phones and tablets, sped up account switching, and made the whole Console feel snappier on every device.

    • Widget Updates Without the Hassle

      Every deploy now delivers the latest SDK to all your customer sites automatically. No embed code changes. No update emails. No missed versions. Just fresh code, every page load.

    • Handoffs That Know When to Step In

      Human handoff now triggers automatically when the AI sees your visitor getting frustrated or hitting dead ends. Location forms prefill from stored context, so your agents start with everything they need from the very first message.

    • A Usage Page You'll Actually Want to Use

      Accurate dates. Proper pagination. Clean layout. The Usage History page is now something you can scan at a glance instead of squinting at.

    • Know Your Visitors at a Glance

      A new Visitor tab in Console conversations shows you who's chatting — profile details right alongside the conversation history.

    • Fresh Look for Your Chat Widget

      New icons, smoother animations, and a design that automatically adapts to your brand colors. We unified the header, moved the copy button so it stops bumping into bubbles, and tightened up the corners. It just feels better.

    Changed4

    • Polished chatbox interactions — The greeting-to-chat transition uses a smooth curtain animation. Input text is vertically centered. Footer spacing is precise. Hero content scrolls on desktop. End Session moved up to the header menu for less clutter.
    • Widget Theme and White Label are now separate — No more confusing overlap between how your widget looks and how your Console is branded. Two distinct features, cleanly handled.
    • Console feels faster — We cut unnecessary network requests from account switching and trimmed overhead behind the scenes. The dashboard is lean and responsive.
    • Smarter embed for your sites — The widget now loads with stricter security guarantees, and loading errors get reported back to your page so they're easier to debug.

    Fixed6

    • Sessions finally stay clean — Stale ended sessions can no longer corrupt new ones. We've tightened how sessions expire so they reliably close instead of lingering.
    • Handoffs work every time — No more duplicate handoffs per session. Visitor profiles sync to integrated help desks once, in the right order. Close-ticket notifications always fire, even when sessions end unexpectedly.
    • Better AI responses — Your chatbot no longer sends empty replies. Period. Internal references to your knowledge sources no longer slip into the answer text. Product and location info stays consistent. Off-topic questions get a graceful "I can't help with that" instead of a confused guess.
    • Billing and credits, sorted — Credit reconciliation works correctly across all tiers. Failed payments retry automatically. Purchased credit packs reconcile properly. Partner accounts always use the partner formula.
    • Chatbox polish — Mobile typing no longer zooms in. Disabled input fields don't steal focus. The copy button plays nice with other bubbles. The greeting screen won't auto-navigate away before you're ready. Account switcher is perfectly centered.
    • Help desk integration improvements — Visitor profiles sync once, in the right order. Ticket timestamps are consistently formatted.

We use cookies to run and improve Gydr.

Read our Cookie Policy