Week 37 v3.7.1
- Add query filters
roomId
andkind
on theuseUnreadInboxNotificationsCount
hook.
v3.7.0
This release introduces group mentions (e.g.
@engineering
) across all packages and first-class support for tenants. Learn more about group mentions and tenants in the docs.-
Add new
resolveGroupsInfo
resolver to provide information about groups (e.g.name
,avatar
, etc) similar toresolveUsers
. -
Support returning group mention suggestions in
resolveMentionSuggestions
. -
Support group mentions in
stringifyCommentBody
, it now accepts aresolveGroupsInfo
option that passes the results to mentions asgroup
. -
Add query filters
roomId
andkind
on thegetInboxNotifications
method. -
Add
useGroupInfo
hook to useresolveGroupsInfo
in React, same asuseUser
forresolveUsers
. -
Add query filters
roomId
andkind
on theuseInboxNotifications
hook. -
Support group mentions in default components (mentions suggestions dropdowns,
Thread
,Composer
,InboxNotification
, etc). -
Support group mentions in text editors and comments-related components.
-
Support group mentions in text editors.
-
Add methods to manage groups on Liveblocks (e.g.
createGroup
,getUserGroups
). -
Add
tenantId
parameters to methods that need it when using tenants. -
Mark
getThreadParticipants
as deprecated, use thread subscriptions orgetMentionsFromCommentBody
instead. -
Support group mentions in
stringifyCommentBody
, it now accepts aresolveGroupsInfo
option that passes the results to mentions asgroup
. -
Support group mentions in email notifications helpers. These functions now accept a
resolveGroupsInfo
option that passes the results to mentions asgroup
.
3.6.2
- Rename
budgetToken
tobudgetTokens
inAnthropicProviderOptions
.
v3.6.1
-
Fixes a bug where a specific combination of concurrent LiveList mutations could break eventual consistency (two clients disagreeing on the final document state).
-
Only show retrieval and reasoning durations in
AiChat
when they are 3 seconds or longer. -
Make
AiTool
titles selectable.
Website
- AI assistant has been added to the documentation and dashboard pages. It can answer
questions on anything related to Liveblocks, such as code, recommendations, bugs,
billing, usage, and project information. It's powered by
AiChat
.
- Add query filters
Week 36 v3.6.0
-
Auto-abort this client's tool calls on page unload to prevent hanging chats.
-
Reasoning in
AiChat
now displays how long it took. -
AiChat
nows shows when a copilot is searching its knowledge defined on the dashboard, as a "Searching 'What is RAG?'…" indicator. It also displays how long it took. -
Add
Duration
primitive to display formatted durations, similar to the existingTimestamp
primitive. -
Better type safety for copilot creation and update options.
-
Add missing type export for AI Copilot and knowledge sources.
v3.5.4
-
Throttle incoming AI delta updates to prevent excessive re-renders during fast streaming.
-
Optimized partial JSON parser for improved tool invocation streaming performance.
-
Fixes a bug where the a comment could not be selected if it was within a previously deleted comment.
Dashboard
- Add API reference modal to AI Copilot detail pages, with React, Node.js, and REST API snippets to get started quickly.
-
Week 35 v3.5.1
- Fixes a bug where deleting a thread/comment from Tiptap would also remove any comments contained within it.
v3.5.0
- Add the following methods for managing AI copilots and knowledge sources:
getAiCopilots
createAiCopilot
getAiCopilot
updateAiCopilot
deleteAiCopilot
createWebKnowledgeSource
createFileKnowledgeSource
deleteFileKnowledgeSource
deleteWebKnowledgeSource
getKnowledgeSources
getKnowledgeSource
getFileKnowledgeSourceMarkdown
getWebKnowledgeSourceLinks
v3.4.2
- Fix improved Markdown streaming in
AiChat
only being enabled in reasoning blocks, it’s now enabled for all Markdown.
v3.4.1
-
Fix a bug where copilot id wasn't passed when setting tool call result if a tool call was defined with
execute
callback. -
Update
useSendAiMessage
to use the the last used copilot id in a chat when no copilot id is passed to the hook or the method returned by the hook.
Website
Week 34 v3.4.0
Tool calls will now stream in while under construction. This means that tools will render sooner and more often re-render, while
partialArgs
are streaming in.New behavior (v3.4 and higher):
- 1st render:
{ stage: "receiving", partialArgs: {} }
- 2nd render:
{ stage: "receiving", partialArgs: { cities: [] } }
- 3rd render:
{ stage: "receiving", partialArgs: { cities: [""] } }
- 4th render:
{ stage: "receiving", partialArgs: { cities: ["Pa"] } }
- 5th render:
{ stage: "receiving", partialArgs: { cities: ["Paris"] } }
- etc.
- Then
{ stage: "executing", args: { cities: "Paris" } }
(same as before) - And
{ stage: "executed", args, result }
(same as before)
Before (v3.3 and lower):
- Stage "receiving" would never happen
- 1st render would be with
{ stage: "executing", args: { cities: ["Paris"] } }
- 2nd render would be with
{ stage: "executed", args, result }
-
In
RoomProvider
,initialPresence
andinitialStorage
now get re-evaluated whenever the room ID (theid
prop) changes. -
Add a minimal appearance to
AiTool
via a newvariant
prop. -
Improve Markdown rendering during streaming in
AiChat
: incomplete content is now handled gracefully so things like bold, links, or tables all render instantly without seeing partial Markdown syntax first. -
Render all messages in
AiChat
as Markdown, including ones from the user. -
Fix Markdown rendering of HTML tags in
AiChat
. (e.g. "Use the<AiChat />
component" would render as "Use the `` component") -
Improve shimmer animation visible on elements like the "Thinking…"/"Reasoning…" placeholders in
AiChat
.
Infrastructure
- Improved LiveList conflict resolution that will keep the conflicting element closer to its intended destination.
- 1st render:
Week 33 v3.3.4
-
Fix race condition where AI tools were not always executing. This could happen when using
useSendAiMessage
first and then immediately opening the<AiChat />
afterwards. -
Scroll thread annotations into view when a thread in
AnchoredThreads
is selected, similarly to@liveblocks/react-lexical
.
v3.3.1
- Fix
Composer
uploading attachments on drop whenshowAttachments
is set tofalse
.
All versions
- Fix attachment names showing URL-encoded characters. (e.g.
a%20file.txt
instead ofa file.txt
)
Infrastructure
- Fixed a bug that caused unreliable storage updates under high concurrency.
- Fixed an issue that could cause LLM responses to appear to "hang" if the token limit got exceeded during the response generation. If this now happens, the response will indicate a clear error to the user.
Dashboard
- New knowledge prompt option when configuring AI copilots, allowing you to customize when back-end knowledge is fetched.
Documentation
- More info on styling AI chat components.
- Disambiguate semantics for
LiveList.delete()
.
-
Week 32 v3.3.0
-
Add
maxVisibleComments
prop toThread
to control the maximum number of comments to show. When comments are hidden, a "Show more replies" button is shown to allow users to expand the thread. -
Add
onComposerSubmit
callback toAiChat
triggered when a new message is sent. It can also be used to customize message submission by callinguseSendAiMessage
yourself. -
Overrides and CSS classes for
AiChat
's composer have been renamed:- Overrides:
AI_CHAT_COMPOSER_SEND
→AI_COMPOSER_PLACEHOLDER
- CSS classes:
.lb-ai-chat-composer-form
→.lb-ai-composer-form
- Overrides:
-
Fix: knowledge passed as a prop to
AiChat
no longer leaks that knowledge to other instances ofAiChat
that are currently mounted on screen. -
Add
query
option touseAiChats
to filter the current user’s AI chats by metadata. Supports exact matches for string values, "contains all" for string arrays, and filtering by absence usingnull
(e.g.{ metadata: { archived: null } }
). -
useSendAiMessage
now accepts passing the chat ID and/or options to the function rather than the hook. This can be useful in dynamic scenarios where the chat ID might not be known when calling the hook for example. -
useCreateAiChat
now accepts a chat ID as a string instead of{ id: "chat-id" }
. -
Allow using custom composers in
FloatingComposer
via thecomponents={{ Composer }}
prop. -
Add
ATTACH_THREAD_COMMAND
command to manually create a thread attached to the current selection.
Dashboard
- Support SAML Single Sign-On for enterprise customers.
- Allow editing first and last name in personal settings.
-
Week 31 v3.2.1
-
Improve Markdown lists in
AiChat
: better spacing and support for arbitrary starting numbers in ordered lists. (e.g.3.
instead of1.
) -
Fix
useSyncStatus
returning incorrect synchronization status for Yjs provider. We now compare the hash of local and remote snapshot to check for synchronization differences between local and remote Yjs document. -
Fix
LiveblocksYjsProvider.getStatus()
returning incorrect synchronization status for Yjs provider.
Dashboard
- Add MAU breakdown to the historical usage table on the “Billing & usage” page (MAU used / Non-billed MAU).
- Support OpenAI compatible AI models in AI Copilots.
- Support Gemini 2.5 Pro and Gemini 2.5 Flash Thinking models in AI Copilots and remove support for the corresponding preview models.
Doocumentation
-
Week 30 v3.2.0
-
Improve
AiChat
's scroll behavior when sending new messages: the chat will now scroll new messages to the top and leave enough space for responses. -
Expose Markdown components in
AiChat
’scomponents
prop to customize the rendering of Markdown content. -
Add
blurOnSubmit
prop toComposer
(also available on theComposer.Form
primitive and asblurComposerOnSubmit
onThread
) to control whether a composer should lose focus after being submitted. -
useErrorListener
now receives"LARGE_MESSAGE_ERROR"
errors when thelargeMessageStrategy
option isn’t configured and a message couldn’t be sent because it was too large for WebSocket. -
Add
tenantId
toidentifyUser
method as an optional parameter.
Dashboard
- Fix issue with custom nodes causing an error in Tiptap/BlockNote documents views.
-
Week 29 v3.1.4
- Fix copilot id not being passed to 'set-tool-call-result' command that is dispatched when a tool call is responded to. Previously, we were using the default copilot to generate messages from the tool call result.
v3.1.3
- Fix
AiChat
component not scrolling instantly to the bottom on render when messages are already loaded.
Dashboard
- Added the ability to use website crawls and sitemaps as knowledge sources for your AI copilot.
Website
- New blog post: What’s new in Liveblocks: June 2025.
Week 28 v3.1.2
- Improve URL sanitization in comments.
v3.1.1
- Adds experimental setting
LiveObject.detectLargeObjects
, which can be enabled globally usingLiveObject.detectLargeObjects = true
(default is false). With this setting enabled, calls toLiveObject.set()
orLiveObject.update()
will throw as soon as you add a value that would make the total size of the LiveObject exceed the platform limit of 128 kB. The benefit is that you get an early error instead of a silent failure, but the downside is that this adds significant runtime overhead if your application makes many LiveObject mutations. - Fix: also display errors in production builds when they happen in
render
methods defined withdefineAiTool()
. Previously, these errors would only be shown during development. - Fix an issue with the render component of tool calls not being displayed correctly when the tool call signal was read before it was registered.
Examples
- New example: AI Support Chat.