• 🎙️ New transcription option: Groq - whisper-large-v3-turbo — added to the Audio Transcription dropdown in Live Chat Settings
• ⚡ Groq's hardware-accelerated inference delivers the fastest Whisper transcription available — typically under 1 second for short audio clips
• 💰 More cost-efficient than OpenAI Whisper for high-volume transcription workloads
• 🔄 Highlighted as the new default in the dropdown (selected option shown in blue)

1. Go to Settings → Live Chat
2. Scroll to the Audio transcription section
3. Select Groq - whisper-large-v3-turbo from the dropdown
4. Save — all incoming audio messages in Live Chat will now be automatically transcribed to text using Groq's Whisper engine

• Groq's LPU (Language Processing Unit) hardware runs Whisper at significantly higher speeds than CPU/GPU-based OpenAI Whisper — reducing transcription latency for agents in real-time support conversations
• Pairs naturally with the Partner Groq API Key setting (released same day) — partners can use their own Groq key to power transcription across all client workspaces
• Enables a fully Groq-powered AI stack within UChat (AI Agents + AI Tasks + Audio Transcription)

👉 Located at: Settings → Live Chat → Audio transcription → Groq - whisper-large-v3-turbo Published: March 10, 2026

• 📎 Batch file selection — select up to 5 images or files in a single send action from the media picker
• 🖼 Multiple selection UI — the media library now shows checkmarks on selected items with a count indicator ("4 / 5 selected")
• 🔵 "Select up to 5" button — clearly indicates the maximum selection limit in the picker toolbar
• 🗂 Library, URL, GIF, and Upload tabs all support the updated multi-select behaviour
• 🗑 Bulk delete — selected items can be deleted in batch using the "Delete Selected" button

1. In Live Chat, open the media/attachment picker
2. Switch to the Library tab (or use Upload Your File, From URL, or GIFs)
3. Click to select up to 5 images or files — a blue checkmark appears on each selected item
4. The toolbar shows the current selection count (e.g., "4 / 5 selected")
5. Click Save — all selected files are sent as attachments in a single message action

• Agents supporting customers who need multiple screenshots, documents, or images no longer need to send them one by one
• Dramatically improves agent efficiency for support tickets involving product photos, invoices, or multi-step visual instructions
• Consistent with multi-file sending behaviour on popular messaging platforms like WhatsApp and Telegram

• 🔐 Verification step added to 2FA setup flow — users must successfully enter the code from their authenticator app before 2FA is marked as enabled
• ✅ Ensures the authenticator app is correctly configured and synced before the feature is active
• 🚫 Prevents the scenario where 2FA is "enabled" but the QR code was scanned incorrectly or the time sync is off — which would result in the user being locked out on next login
• 🔄 No change to the 2FA login flow — only the initial setup/enable step is affected

1. Go to Profile → Security
2. Enable Two-Factor Authentication — a QR code is displayed
3. Scan the QR code with your authenticator app (Google Authenticator, Authy, etc.)
4. NEW: Enter the 6-digit code from your authenticator app to confirm it is working correctly
5. 2FA is now activated — only after successful verification

• Eliminates the risk of users enabling 2FA with a broken/unsynced authenticator setup and then losing access to their account
• Follows security best practices — all major platforms (Google, GitHub, Stripe) require verification before 2FA activation
• Reduces support tickets related to 2FA lockouts caused by misconfigured authenticator apps

• 🤖 List AI Agents via API — retrieve all configured AI Agents in the workspace with a single GET call
• ⚙️ Update AI Agent provider & model via API — programmatically switch the AI provider (OpenAI, Groq, Anthropic, xAI, etc.) and specific model for any AI Agent
• 🔑 All endpoints are authenticated (lock icon visible) — require a valid API token
• 🔄 Consistent with the parallel AI Tasks endpoints (/flow/ai-tasks and /flow/update-ai-task-provider) for a unified AI Hub API pattern

1. Authenticate with your API key
2. Call GET /flow/ai-agents to retrieve the list of AI Agents and their current provider/model assignments
3. Call POST /flow/update-ai-agent-provider with the Agent ID and the desired provider + model in the request body to update the configuration

• Enables automated AI model management across all flows — switch providers or upgrade to newer models without touching the UI
• Useful for partners provisioning new workspaces with standardised AI configurations
• Pairs with the Default AI Provider setting in AI Hub Settings (released same week) for end-to-end AI configuration automation

👉 Located at: API Docs → Flow AI Hub Published: February 14, 2026

{
  "data": [
    {
      "user_id": 0,
      "workspace_id": 0,
      "workspace_name": "string",
      "role": "string"
    }
  ]
}

{ "message": "Error message" }

• 🔍 Search for any user by email via Partner API — look up account details and workspace memberships in one call
• 📋 Returns user_id, workspace_id, workspace_name, and role for every workspace the user belongs to
• 📄 Supports pagination via limit and page parameters (up to page 100)
• 🔑 Requires Manage Team scope on your Partner API token

1. Authenticate with your Partner API key (Manage Team scope required)
2. Call GET /partner/find-user-by-email?email=user@example.com
3. Optionally add &limit=20&page=1 for paginated results
4. Parse the data array to see which workspaces the user is a member of and their role in each

• Enables partners to audit user workspace access across all client accounts programmatically
• Useful for support workflows — quickly locate which workspaces a user belongs to by email without manual UI searching
• Supports account management automation, such as cross-workspace user deprovisioning or role auditing

👉 Partner-only feature. Requires Manage Team scope. Published: February 12, 2026

• 🔢 New field: Widget Code — a unique, editable code for your widget (e.g., 95-yourbrandname) added to the Edit Widget screen
• 🔗 The Widget Code is used directly in the Install JavaScript Snippet URL, replacing the default numeric ID
• ✅ Must be unique within the workspace — a validation hint is shown below the field
• 📋 Install snippet updates automatically when a Widget Code is set — copy it directly from the widget editor

<!-- With Widget Code -->
<script src="https://www.uchat.com.au/widgetf/{WidgetCode}" async="async"></script>

<!-- Example -->
<script src="https://www.uchat.com.au/widgetf/12416w/144207.js" async="async"></script>

1. Go to Mini-Apps → Widgets and open or create a Chat Button type widget
2. In the Widget Code field (highlighted in red, left panel), enter a unique code for your widget
3. The code must be unique within your workspace
4. Copy the updated Install JavaScript Snippet from the bottom of the left panel
5. Paste the snippet into the <head> of your website pages where you want the chat button to appear

• Gives partners and whitelabel users human-readable, branded widget URLs instead of opaque numeric IDs
• Makes it easier to manage and identify widgets across multiple client websites
• Simplifies widget deployment scripts when managing many sites — recognisable codes are less error-prone than long numeric IDs

👉 Located at: Mini-Apps → Widgets → Edit Widget → Widget Code Published: February 9, 2026

• 🇷🇺 New language: Russian (ru) — added to the Language selector in Profile settings
• 🌐 Full UI translation covering flow builder, live chat, analytics, and workspace settings
• ✅ Selected and confirmed as the active language — highlighted in the dropdown when chosen

1. Click your profile avatar (top-right) and go to Profile
2. Under Contact Information, find the Language dropdown
3. Select Russian (ru) from the list
4. The UI will immediately switch to Russian

• 🔇 New question item: Silent Input — adds a waiting step that listens for user input without sending any outgoing message
• 👁 Displayed with a hidden/eye-off icon (🚫👁) in the flow canvas to distinguish it from regular question items
• 💾 Supports Save Response to field — capture the silent input into any custom user field
• 🔀 Includes No Match and No Input exit paths — handle invalid responses or timeouts just like standard question items
• ⚙️ Compatible with Advanced Settings and Button Click Handler for full flow control
• 📱 Works across all supported channels (WeChat, Messenger, Instagram, Telegram, TikTok, WhatsApp, Line, Viber, VK, etc.)

1. Open or create a Flow in Flow Builder
2. Add a Question node to your canvas
3. Click + Add Item inside the Question node
4. Select Silent Input from the item type list
5. (Optional) Set a Save Response to field to capture whatever the user sends
6. Connect the No Match, No Input, and Continue to next step exit paths as needed

• Capture a user's free-text reply after a previous message node without repeating a question
• Wait silently for an inbound message at a specific point in the flow (e.g., awaiting a payment confirmation or image upload)
• Build multi-step input flows where the prompt is delivered by a preceding Message node and the Question node only needs to capture the response

• Separates the "display a message" concern from the "capture user input" conce

• 🏢 New integration: Google My Business — found under Integrations → Others → Google My Business
• 🔗 Connect your Google account via OAuth to link your GMB profile
• 📋 List reviews by location — retrieve Google reviews for any of your business locations
• 💬 Reply to reviews by location — send replies to customer reviews directly from a flow
• 🔄 Reconnect support — easily reauthorise your Google account if the connection expires

1. Go to Integrations in the left sidebar
2. Scroll to the Others section and select Google My Business
3. Click Connect and authorise with your Google account
4. Your linked business location will appear in the integration panel
5. In your Flow builder, use the Google My Business actions to List Reviews or Reply to a Review for a selected location

• Automatically monitor and respond to new Google reviews as part of a customer feedback flow
• Trigger follow-up messages or internal alerts when a review is received
• Build reputation management automations that reply to reviews based on star rating or keywords

• Brings Google review management into UChat's automation ecosystem — no need to switch between platforms
• Especially valuable for local service businesses (e.g., ProfitsLocal clients) managing multiple Google Business locations
• Pairs naturally with Google Map integration already available in UChat

👉 Located at: Integrations → Others → Google My Business Published: November 19, 2025

{
  "chat": 1,
  "facebook": 1,
  "instagram": 1,
  "telegram": 1,
  "slack": 1,
  "whatsapp": 1,
  "whatsapp_cloud": 1,
  "wechat": 1,
  "voice": 1,
  "sms": 1,
  "rcs": 1,
  "line": 1,
  "viber": 1,
  "vk": 1,
  "intercom": 1,
  "jivochat": 1,
  "chatwoot": 1,
  "tiktok": 1
}

• 🔌 Control channel visibility via API — show or hide any channel in the workspace sidebar programmatically
• ⚙️ No parameters required — all configuration is passed in the request body
• 🔑 Requires Manage Team scope on your API token
• 📦 Response mirrors the request body, confirming the updated channel visibility state

1. Authenticate with your API key (Manage Team scope required)
2. Call POST /workspace-settings/update-channels
3. Pass a JSON body with the channels you want to show (1) or hide (0)
4. The 200 response returns the updated data object confirming each channel's new state

{
  "data": [
    {
      "channel": "facebook",
      "is_linked": true,
      "linked_name": "string"
    }
  ],
  "status": "ok"
}

• 🤖 Retrieve Omni Bot channel links via API — check which channels (Facebook, WhatsApp, Instagram, Telegram, etc.) are connected to the Omni Bot in any workspace
• 🔍 Returns channel, is_linked status, and linked_name for each channel
• 🔑 Requires Manage Team scope on your Partner API token

1. Authenticate with your Partner API key (Manage Team scope required)
2. Call GET /partner/workspace/{workspaceId}/get-omni-bot-linked-channels with the target workspace ID
3. Parse the data array to see which channels are currently linked and their display names

• Enables partners to audit Omni Bot channel configurations across all client workspaces programmatically
• Useful for onboarding workflows — verify which channels are live before triggering automated flows
• Supports partner dashboards that need to display channel status per workspace without requiring UI access

👉 Partner-only feature. Requires Manage Team scope. Published: November 16, 2025.