# assistant-ui

> React components for AI chat interfaces

## LLM Documentation Files

- [Full documentation](https://www.assistant-ui.com/llms-full.txt): every documentation page rendered into one large text file.
- Per-page markdown: append `.md` to any docs page URL. `.mdx` is kept as a backwards-compatible alias for agents that request source-style URLs. For example, `/docs/installation.md` and `/docs/installation.mdx` both return markdown for `/docs/installation`.
- Markdown by Accept header: requesting a docs, examples, design or elements page with `Accept: text/markdown` also returns that page's markdown.
- Use the index below to choose a specific page. Remove the `.md` or `.mdx` suffix to open the human-readable docs page.

## Agent Discovery

- [Agent instructions](https://www.assistant-ui.com/AGENTS.md)
- [Site skill](https://www.assistant-ui.com/skill.md)
- [API catalog](https://www.assistant-ui.com/.well-known/api-catalog)
- [Agent Skills index](https://www.assistant-ui.com/.well-known/agent-skills/index.json)
- Agent skills: task-shaped SKILL.md guides listed in the Agent Skills index, one per area (assistant-ui, cloud, copilots, elements, generative-ui, ink, markdown, observability, primitives, react-mcp, react-native, runtime, setup, streaming, thread-list, tools, update).
- [Markdown sitemap](https://www.assistant-ui.com/sitemap.md)
- [Documentation MCP endpoint](https://www.assistant-ui.com/mcp)

## Table of Contents

### architecture

- [Architecture](https://www.assistant-ui.com/docs/architecture.md): How components, runtimes, and cloud services fit together.

### root

- [Overview](https://www.assistant-ui.com/docs.md): Components, runtimes, and primitives for building AI chat interfaces in React, Vue, React Native, and the terminal.

### llm

- [Agent Skills](https://www.assistant-ui.com/docs/llm.md): Use AI tools to build with assistant-ui faster. AI-accessible documentation, Claude Code skills, and MCP integration.

### base-ui

- [Radix UI and Base UI](https://www.assistant-ui.com/docs/base-ui.md): How the assistant-ui component registry serves Radix UI and Base UI flavored components for every shadcn style.

### cli

- [CLI](https://www.assistant-ui.com/docs/cli.md): Scaffold projects, add components, and manage updates from the command line.

### devtools

- [DevTools](https://www.assistant-ui.com/docs/devtools.md): Inspect runtime state, context, and events in the browser.

### installation

- [Installation](https://www.assistant-ui.com/docs/installation.md): Get assistant-ui running in 5 minutes with npm and your first chat component.

### rtl

- [RTL Support](https://www.assistant-ui.com/docs/rtl.md): Use assistant-ui with right-to-left languages like Arabic, Hebrew, and Persian.

### api-reference

- [Attachment Adapters](https://www.assistant-ui.com/docs/api-reference/adapters/attachments.md): Attachment adapters for uploading files, handling lifecycle events, and bringing app-owned content into assistant-ui com
- [Feedback Adapter](https://www.assistant-ui.com/docs/api-reference/adapters/feedback.md): Capture and respond to message feedback submitted through action primitives or runtime actions.
- [Adapters API Reference](https://www.assistant-ui.com/docs/api-reference/adapters.md): Adapter interfaces for connecting chat models, persistence, file attachments, feedback, and suggestions to assistant-ui 
- [Model Adapters](https://www.assistant-ui.com/docs/api-reference/adapters/model.md): Adapter interfaces for connecting chat models, streaming responses, and model execution to assistant-ui runtimes.
- [Persistence Adapters](https://www.assistant-ui.com/docs/api-reference/adapters/persistence.md): Persistence adapters for saving assistant-ui message history, remote thread lists, and long-running chat sessions across
- [Runtime Adapter Context](https://www.assistant-ui.com/docs/api-reference/adapters/runtime.md): Provide assistant-ui runtime adapters through React context for model, attachment, speech, and feedback behavior.
- [Suggestion Adapters](https://www.assistant-ui.com/docs/api-reference/adapters/suggestions.md): Suggestion adapters for providing starter prompts, contextual actions, and guided composer options to assistant-ui runti
- [AssistantRuntimeProvider](https://www.assistant-ui.com/docs/api-reference/context-providers/assistant-runtime-provider.md): Root React provider that connects an assistant-ui runtime to primitives, hooks, threads, and composer state.
- [Context Providers API Reference](https://www.assistant-ui.com/docs/api-reference/context-providers.md): React context providers including AssistantRuntimeProvider that scope assistant-ui runtime, thread, message part, and at
- [Scoped Providers](https://www.assistant-ui.com/docs/api-reference/context-providers/scoped-providers.md): Lower-level assistant-ui providers for custom renderers, scoped message parts, attachments, and advanced composition.
- [External Store API Reference](https://www.assistant-ui.com/docs/api-reference/external-store.md): External store runtime, message conversion helpers, and adapters for assistant-ui React apps that own their chat state o
- [Message Conversion](https://www.assistant-ui.com/docs/api-reference/external-store/message-conversion.md): Convert external message formats into assistant-ui's message and thread state for the external store runtime.
- [External Store Runtime](https://www.assistant-ui.com/docs/api-reference/external-store/runtime.md): Runtime components, options, and adapters for using assistant-ui with externally owned chat state.
- [A2UI](https://www.assistant-ui.com/docs/api-reference/generative-ui/a2ui.md): Convert A2UI surface operations into generative UI specs.
- [Generative UI Actions](https://www.assistant-ui.com/docs/api-reference/generative-ui/actions.md): Register handlers for model-emitted actions and dispatch them from interactive generative UI components.
- [Generative UI Components](https://www.assistant-ui.com/docs/api-reference/generative-ui/components.md): Define the component library JSONGenerativeUI can render, including schemas, render functions, and the default vocabular
- [Generative UI API Reference](https://www.assistant-ui.com/docs/api-reference/generative-ui.md): Spec-driven generative UI for assistant-ui. The data format an assistant streams, the component registry that resolves i
- [JSONGenerativeUI](https://www.assistant-ui.com/docs/api-reference/generative-ui/json-generative-ui.md): Create present and prompt_user tools from a generative UI component library, with options for display mode and action di
- [Generative UI Rendering](https://www.assistant-ui.com/docs/api-reference/generative-ui/rendering.md): Render generative UI trees, build present-tool schemas, and inspect or serialize model-produced UI nodes.
- [Slack Block Kit](https://www.assistant-ui.com/docs/api-reference/generative-ui/slack.md): Convert generative UI trees to Slack Block Kit payloads and decode block_actions interactions back into $action payloads
- [Generative UI Spec](https://www.assistant-ui.com/docs/api-reference/generative-ui/spec.md): The serializable node tree an assistant emits to describe generative UI. Covers the GenerativeUISpec format, its nodes, 
- [Microsoft Teams](https://www.assistant-ui.com/docs/api-reference/generative-ui/teams.md): Convert generative UI trees to Teams Adaptive Cards and decode Action.Submit payloads back into $action payloads.
- [Generative UI Tokens](https://www.assistant-ui.com/docs/api-reference/generative-ui/tokens.md): Shared token arrays used by the default generative UI vocabulary for sizing, color, alignment, and component variants.
- [Composer Trigger Hooks](https://www.assistant-ui.com/docs/api-reference/hooks/composer-triggers.md): Unstable assistant-ui hooks for mention menus, slash commands, and custom composer trigger popovers.
- [Hooks API Reference](https://www.assistant-ui.com/docs/api-reference/hooks.md): React hooks for assistant-ui: useAui, useAuiState, runtime creation, model context registration, and helpers for buildin
- [Model Context Hooks](https://www.assistant-ui.com/docs/api-reference/hooks/model-context.md): React hooks for registering assistant-ui tools, data renderers, instructions, and model context providers.
- [Primitive Hooks](https://www.assistant-ui.com/docs/api-reference/hooks/primitives.md): Primitive hooks for reading scoped assistant-ui runtime state, viewport behavior, timing, and message part data inside R
- [Runtime Hooks](https://www.assistant-ui.com/docs/api-reference/hooks/runtimes.md): Runtime creation hooks for local, remote, cloud, external-store, and AI SDK powered assistant-ui chat experiences.
- [State Hooks](https://www.assistant-ui.com/docs/api-reference/hooks/state.md): State selector and action hooks for reading assistant-ui runtime state and controlling threads, composers, messages, and
- [@assistant-ui/ai-sdk](https://www.assistant-ui.com/docs/api-reference/integrations/ai-sdk.md): Vercel AI SDK runtime hooks, chat transports, and message conversion utilities for assistant-ui applications.
- [assistant-cloud/ai-sdk](https://www.assistant-ui.com/docs/api-reference/integrations/assistant-cloud-ai-sdk.md): The AI SDK message format adapter and run telemetry extraction that the assistant-ui runtime uses, typed against UIMessa
- [assistant-cloud/telemetry](https://www.assistant-ui.com/docs/api-reference/integrations/assistant-cloud-telemetry.md): OpenTelemetry exporter, span processor and stream metadata helpers that send a server's GenAI spans to Assistant Cloud a
- [assistant-cloud](https://www.assistant-ui.com/docs/api-reference/integrations/assistant-cloud.md): The assistant-cloud client for threads, messages, runs, events, scores and files, with the reporters, persistence and ru
- [@assistant-ui/eve](https://www.assistant-ui.com/docs/api-reference/integrations/eve.md): Eve runtime hook and message conversion utilities for assistant-ui React applications.
- [Integrations API Reference](https://www.assistant-ui.com/docs/api-reference/integrations.md): Package-level APIs for connecting assistant-ui React to the Vercel AI SDK, Assistant Cloud, and adjacent chat ecosystem 
- [@assistant-ui/react-data-stream](https://www.assistant-ui.com/docs/api-reference/integrations/react-data-stream.md): Data stream runtime hook and message conversion utilities for custom assistant-ui streaming backends.
- [Model Context](https://www.assistant-ui.com/docs/api-reference/model-context/context.md): Provide model instructions, contextual state, and inline renderers to assistant-ui runtimes.
- [Model Context API Reference](https://www.assistant-ui.com/docs/api-reference/model-context.md): Model instructions, contextual state, provider registries, and renderers for giving assistant-ui runtimes app-aware cont
- [Model Context Registry](https://www.assistant-ui.com/docs/api-reference/model-context/registry.md): Register and manage assistant-ui model context providers that contribute instructions and app state.
- [API Reference](https://www.assistant-ui.com/docs/api-reference/overview.md): Complete assistant-ui React API reference for building AI chat UIs with primitives, hooks, runtimes, adapters, tools, tr
- [ActionBarMorePrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/action-bar-more.md): Overflow menu primitives for grouping secondary assistant message actions in a custom React UI.
- [ActionBarPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/action-bar.md): Composable message action controls for copy, edit, reload, speech, and feedback in assistant-ui chat interfaces.
- [AuiIf](https://www.assistant-ui.com/docs/api-reference/primitives/assistant-if.md): Conditional rendering primitive for showing React UI from assistant-ui thread, message, composer, and runtime state.
- [AssistantModalPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/assistant-modal.md): Floating assistant modal primitives for building support chat, copilot, and embedded assistant experiences.
- [AttachmentPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/attachment.md): Attachment primitives for rendering file previews, names, thumbnails, and remove controls in assistant-ui messages and c
- [BranchPickerPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/branch-picker.md): Branch picker primitives for navigating regenerated assistant responses and alternate message paths inside a chat thread
- [ChainOfThoughtPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/chain-of-thought.md): Chain of thought primitives for rendering assistant reasoning, step lists, and collapsible disclosure UI in message cont
- [ComposerPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/composer.md): Composable input primitives for assistant-ui prompts, send controls, cancellation, attachments, and composer state.
- [Composition](https://www.assistant-ui.com/docs/api-reference/primitives/composition.md): How to compose primitives with custom components using asChild.
- [ErrorPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/error.md): Error primitives for rendering assistant-ui runtime, thread, and message failures inside custom chat interfaces.
- [Primitives API Reference](https://www.assistant-ui.com/docs/api-reference/primitives.md): Composable React primitives for assistant-ui chat UIs: Thread, Composer, Message, BranchPicker, ActionBar, and the parts
- [MessagePartPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/message-part.md): Message part primitives for rendering text, tool calls, data parts, reasoning, source content, and custom assistant outp
- [MessagePrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/message.md): Message primitives for rendering assistant and user turns, message parts, attachments, actions, editing, and branch cont
- [QueueItemPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/queue-item.md): Queue item primitives for rendering pending assistant-ui thread operations, optimistic work, and runtime queue state.
- [SelectionToolbarPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/selection-toolbar.md): Selection toolbar primitives for quote, copy, and contextual actions on selected chat text.
- [SuggestionPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/suggestion.md): Suggestion primitives for rendering starter prompts, follow-up actions, and composer suggestions in assistant-ui threads
- [ThreadListItemMorePrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/thread-list-item-more.md): Overflow menu primitives for secondary thread list item actions in custom assistant-ui sidebars.
- [ThreadListItemPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/thread-list-item.md): Thread list item primitives for rendering selectable conversation rows with titles, archive controls, delete actions, an
- [ThreadListPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/thread-list.md): Thread list primitives for rendering conversation navigation, new thread actions, and custom assistant sidebars.
- [ThreadPrimitive](https://www.assistant-ui.com/docs/api-reference/primitives/thread.md): Thread primitives for rendering chat transcripts, message lists, viewport state, suggestions, and composers in assistant
- [AssistantRuntime](https://www.assistant-ui.com/docs/api-reference/runtimes/assistant-runtime.md): Top-level assistant-ui runtime actions and state for tools, threads, composers, messages, and assistant behavior.
- [AttachmentRuntime](https://www.assistant-ui.com/docs/api-reference/runtimes/attachment-runtime.md): AttachmentRuntime state and actions for reading attachment data and controlling files inside assistant-ui messages and c
- [ComposerRuntime](https://www.assistant-ui.com/docs/api-reference/runtimes/composer-runtime.md): ComposerRuntime state and actions for controlling assistant-ui composer text, attachments, submission, cancellation, and
- [Runtime State API Reference](https://www.assistant-ui.com/docs/api-reference/runtimes.md): Runtime state and actions exposed through useAui and useAuiState, covering AssistantRuntime, ThreadRuntime, ThreadListRu
- [MessagePartRuntime](https://www.assistant-ui.com/docs/api-reference/runtimes/message-part-runtime.md): MessagePartRuntime state and helpers for inspecting assistant-ui text, tool calls, data parts, reasoning, and custom mes
- [MessageRuntime](https://www.assistant-ui.com/docs/api-reference/runtimes/message-runtime.md): MessageRuntime state and actions for editing, reloading, copying, rating, speaking, and branching assistant-ui messages.
- [QueueItemState](https://www.assistant-ui.com/docs/api-reference/runtimes/queue-state.md): State shape for queued assistant-ui thread operations and pending runtime work.
- [ThreadListItemRuntime](https://www.assistant-ui.com/docs/api-reference/runtimes/thread-list-item-runtime.md): ThreadListItemRuntime state and actions for selecting, archiving, unarchiving, deleting, and renaming assistant-ui conve
- [ThreadListRuntime](https://www.assistant-ui.com/docs/api-reference/runtimes/thread-list-runtime.md): ThreadListRuntime state and actions for managing remote assistant-ui conversations, active thread selection, and new thr
- [ThreadRuntime](https://www.assistant-ui.com/docs/api-reference/runtimes/thread-runtime.md): ThreadRuntime state and actions for controlling assistant-ui messages, composers, suggestions, model context, and the fu
- [Component Tools](https://www.assistant-ui.com/docs/api-reference/tools/component-tools.md): Register assistant tools from mounted React components, scoped to the lifetime of part of the UI tree.
- [Tools API Reference](https://www.assistant-ui.com/docs/api-reference/tools.md): Tool definitions, React renderers, status helpers, and toolkits for exposing callable app capabilities to assistant-ui c
- [Interactables (legacy)](https://www.assistant-ui.com/docs/api-reference/tools/interactables-legacy.md): Deprecated legacy interactables APIs for registering model-editable components with per-instance update tools.
- [Interactables](https://www.assistant-ui.com/docs/api-reference/tools/interactables.md): Unstable interactables APIs for model-editable app and message state, including hooks, resources, toolkit helpers, and s
- [Tool Rendering](https://www.assistant-ui.com/docs/api-reference/tools/rendering.md): Register React renderers for assistant-ui tool calls, tool results, and model data parts.
- [Tool Status](https://www.assistant-ui.com/docs/api-reference/tools/status.md): Read tool arguments, execution status, and result state inside assistant-ui tool UI components.
- [Toolkits](https://www.assistant-ui.com/docs/api-reference/tools/toolkits.md): Define model-facing tools and compose them into named toolkits registered with an assistant-ui runtime scope.
- [Assistant Transport](https://www.assistant-ui.com/docs/api-reference/transport/assistant-transport.md): Command, protocol, and transport types for connecting assistant-ui runtimes across execution boundaries.
- [Assistant Frame](https://www.assistant-ui.com/docs/api-reference/transport/frame.md): Frame bridge APIs and serialized message types for embedding assistant-ui runtimes in external contexts.
- [Transport API Reference](https://www.assistant-ui.com/docs/api-reference/transport.md): Transport commands, frame messages, and protocol types for synchronizing assistant-ui runtimes across process or iframe 
- [Utilities API Reference](https://www.assistant-ui.com/docs/api-reference/utilities.md): Utility exports for custom rendering, composition, and advanced assistant-ui behavior that does not fit a larger API fam
- [Utilities](https://www.assistant-ui.com/docs/api-reference/utilities/miscellaneous.md): Miscellaneous @assistant-ui/react utilities for custom rendering, composition, and advanced assistant UI behavior.
- [Voice API Reference](https://www.assistant-ui.com/docs/api-reference/voice.md): Realtime voice, speech synthesis, and dictation contracts for wiring spoken assistant flows into React chat UIs.
- [Voice Sessions](https://www.assistant-ui.com/docs/api-reference/voice/session.md): Create and control realtime assistant-ui voice sessions, state, controls, and helpers.
- [Speech and Dictation](https://www.assistant-ui.com/docs/api-reference/voice/speech-dictation.md): Connect speech synthesis and dictation adapters to assistant-ui voice and composer workflows.

### cloud

- [Allowed origins](https://www.assistant-ui.com/docs/cloud/allowed-origins.md): Control which browser origins can reach a project's frontend API host through CORS.
- [Anonymous sessions](https://www.assistant-ui.com/docs/cloud/anonymous-sessions.md): Give a browser visitor a cloud identity before sign in, keep its threads for 30 days, and claim them after authenticatio
- [API keys](https://www.assistant-ui.com/docs/cloud/api-keys.md): Create, use, monitor, and revoke a server side API key for Assistant Cloud.
- [Auth providers](https://www.assistant-ui.com/docs/cloud/auth-providers.md): Trust external JWT issuers with auth rules, exchange their tokens in the browser, and mint cloud tokens from a server wh
- [Authentication](https://www.assistant-ui.com/docs/cloud/authorization.md): Choose an Assistant Cloud client mode, follow a request through authentication, and understand the errors the API return
- [Users and workspaces](https://www.assistant-ui.com/docs/cloud/users-and-workspaces.md): Decide how Assistant Cloud identifies a user, choose the workspace that owns their threads, and manage user visibility a
- [Attachments](https://www.assistant-ui.com/docs/cloud/attachments.md): Upload images and documents from the composer, send them as message parts, and read them through short lived signed URLs
- [Messages](https://www.assistant-ui.com/docs/cloud/messages.md): How Assistant Cloud stores conversation trees, converts message formats, persists runtime history and records message fe
- [Retention](https://www.assistant-ui.com/docs/cloud/retention.md): Choose how long Assistant Cloud keeps conversation data, schedule shorter windows safely, and understand what remains fo
- [Thread titles](https://www.assistant-ui.com/docs/cloud/thread-titles.md): A model names each conversation that has a user message, from that message. How a title is made, every setting behind it
- [Threads](https://www.assistant-ui.com/docs/cloud/threads.md): How an app creates, loads, updates, archives and deletes cloud conversations, and how those conversations appear in the 
- [Assistants](https://www.assistant-ui.com/docs/cloud/assistants.md): Configure a stored provider and model, then run it from a cloud thread.
- [Evaluators](https://www.assistant-ui.com/docs/cloud/evaluators.md): Define model judges that write structured verdicts for sampled conversations.
- [Harnesses](https://www.assistant-ui.com/docs/cloud/harnesses.md): Run an AI SDK endpoint at a dedicated cloud origin while keeping conversations in Assistant Cloud.
- [Intelligence](https://www.assistant-ui.com/docs/cloud/intelligence.md): Classify conversations into topics, tasks, questions, and signals with a model you choose.
- [LLM providers](https://www.assistant-ui.com/docs/cloud/llm-providers.md): Connect the model accounts that Assistant Cloud features use for generation, analysis, and evaluation.
- [AI SDK](https://www.assistant-ui.com/docs/cloud/ai-sdk.md): Connect the AI SDK runtime to Assistant Cloud for persisted threads, titles, run reports, engagement, feedback and attac
- [Custom thread list](https://www.assistant-ui.com/docs/cloud/custom-thread-list.md): Compose a Cloud backed thread list around a runtime, adapter or message format of your own.
- [LangGraph, LangChain and ADK](https://www.assistant-ui.com/docs/cloud/langgraph.md): Add an Assistant Cloud thread list, a copy of the conversation, feedback and engagement events to runtimes whose backend
- [Local runtime](https://www.assistant-ui.com/docs/cloud/local-runtime.md): Give any backend Assistant Cloud threads, persistence, titles and run reports through useLocalRuntime, or compose the cl
- [Migrate from Cloud AI SDK](https://www.assistant-ui.com/docs/cloud/migrate-cloud-ai-sdk.md): @assistant-ui/cloud-ai-sdk, the Cloud AI SDK, is deprecated. The last published version keeps working; new work moves to
- [Which runtime](https://www.assistant-ui.com/docs/cloud/runtimes.md): What every assistant-ui runtime stores in the cloud and reports to it, what each cannot, and how the three platforms dif
- [Servers and bots](https://www.assistant-ui.com/docs/cloud/servers.md): Store bot conversations, report model work and read an Assistant Cloud project from a backend with an API key.
- [Stored message formats](https://www.assistant-ui.com/docs/cloud/formats.md): The message formats Assistant Cloud stores, converts, and reads for thread titles.
- [Limits](https://www.assistant-ui.com/docs/cloud/limits.md): Every limit a project enforces, in one place.
- [The assistant-cloud client](https://www.assistant-ui.com/docs/cloud/sdk.md): Construct the Assistant Cloud client, choose authentication, and use its namespaces.
- [Engagement events](https://www.assistant-ui.com/docs/cloud/engagement.md): Record the actions users take around an answer, how events reach Assistant Cloud, and how the dashboard turns them into 
- [Run reports](https://www.assistant-ui.com/docs/cloud/run-reports.md): The report sent for an assistant response, its limits and outcomes, and how it becomes a run in Assistant Cloud.
- [Feedback and scores](https://www.assistant-ui.com/docs/cloud/scores.md): Collect message thumbs and write, read, and inspect named scores for threads, messages, and runs.
- [Traces](https://www.assistant-ui.com/docs/cloud/traces.md): Send server model and tool spans to Assistant Cloud, map them to runs, and merge them with a browser report.
- [Auth tokens](https://www.assistant-ui.com/docs/cloud/api/auth-tokens.md): Mint short lived access tokens for trusted backends and anonymous visitors.
- [Events](https://www.assistant-ui.com/docs/cloud/api/events.md): Record text-free engagement events and update a project's daily rollups and thread signals.
- [Files](https://www.assistant-ui.com/docs/cloud/api/files.md): Generate signed upload and download URLs for project attachments.
- [Conventions](https://www.assistant-ui.com/docs/cloud/api.md): The hosts, credentials, headers, ids, paging rules and error shapes every route of the project API shares, and the index
- [MCP](https://www.assistant-ui.com/docs/cloud/api/mcp.md): Connect an MCP client to Assistant Cloud project reads.
- [Messages](https://www.assistant-ui.com/docs/cloud/api/messages.md): Create, read, delete, update, and score the messages in a thread.
- [Project read API](https://www.assistant-ui.com/docs/cloud/api/project.md): Read project runs, usage, threads and scores from a backend with an API key.
- [Runs](https://www.assistant-ui.com/docs/cloud/api/runs.md): Record completed, incomplete, or failed model runs and start streamed assistant runs.
- [Scores](https://www.assistant-ui.com/docs/cloud/api/scores.md): Record numeric, categorical and boolean scores against a thread, message or run.
- [Threads](https://www.assistant-ui.com/docs/cloud/api/threads.md): Create, list, read, update, archive and delete threads, and claim an anonymous visitor's threads after sign in.
- [Traces](https://www.assistant-ui.com/docs/cloud/api/traces.md): Receive OTLP traces and map GenAI spans into runs, steps and tool calls.
- [Users](https://www.assistant-ui.com/docs/cloud/api/users.md): Erase a user's project data through the project wide API.
- [Concepts](https://www.assistant-ui.com/docs/cloud/concepts.md): The objects a project holds, how a request names the user and the workspace it acts for, and what the SDK sends during o
- [Conversation views](https://www.assistant-ui.com/docs/cloud/dashboard/conversation-views.md): How the thread page draws what each runtime stores, and how to show a conversation with your own components.
- [Engagement page](https://www.assistant-ui.com/docs/cloud/dashboard/engagement.md): See what people do around answers, from first replies and second turns to reactions, features, and satisfaction signals.
- [Exports](https://www.assistant-ui.com/docs/cloud/dashboard/exports.md): Download the current Threads, Runs, or Users view as CSV or JSONL.
- [Using the dashboard](https://www.assistant-ui.com/docs/cloud/dashboard.md): How the dashboard at cloud.assistant-ui.com is laid out, how ranges, filters and time work on every page, who may change
- [Intelligence page](https://www.assistant-ui.com/docs/cloud/dashboard/intelligence.md): Read the topics, tasks, questions, signals, and judged outcomes from Assistant Cloud analysis.
- [Models page](https://www.assistant-ui.com/docs/cloud/dashboard/models.md): Read model cost, usage, cache share, latency, and outcomes by model and provider.
- [Overview page](https://www.assistant-ui.com/docs/cloud/dashboard/overview.md): Read the project's range of runs, people, cost, outcomes, models and recent conversations from one page.
- [Runs page](https://www.assistant-ui.com/docs/cloud/dashboard/runs.md): Filter every response report, compare latency and outcomes, and inspect the run, trace and spans behind one response.
- [Threads page](https://www.assistant-ui.com/docs/cloud/dashboard/threads.md): Search every conversation, archived ones included, filter it by behaviour or analysis, and inspect its transcript, runs 
- [Users page](https://www.assistant-ui.com/docs/cloud/dashboard/users.md): Find the people who use the assistant, how often they return, and the threads and runs behind their activity.
- [Introduction](https://www.assistant-ui.com/docs/cloud.md): Assistant Cloud persists conversations, reports every run, records what users do, and gives you a dashboard over threads
- [Plans and pricing](https://www.assistant-ui.com/docs/cloud/pricing.md): What each plan includes, how active users are counted, and what happens at the cap.
- [Quickstart](https://www.assistant-ui.com/docs/cloud/quickstart.md): From an empty project to a persisted, reported and titled conversation in the dashboard, and what to check when the firs
- [Alerts](https://www.assistant-ui.com/docs/cloud/settings/alerts.md): Watch a project metric once an hour and receive a signed webhook when it is above a threshold.
- [Audit log](https://www.assistant-ui.com/docs/cloud/settings/audit-log.md): See the configuration and data changes made in a project, who made them and the fields that changed.
- [Billing and usage](https://www.assistant-ui.com/docs/cloud/settings/billing.md): Read the current plan, active user allowance and usage projection, then break the billing period down by day, model and 
- [General](https://www.assistant-ui.com/docs/cloud/settings/general.md): Project identity, data retention, and support managed project changes.
- [Settings](https://www.assistant-ui.com/docs/cloud/settings.md): Every settings page of a project, where each is documented, who may change it, and what the audit log records.
- [Model prices](https://www.assistant-ui.com/docs/cloud/settings/model-prices.md): Override model prices for stored runs and find model IDs the catalog cannot price.
- [Telemetry settings](https://www.assistant-ui.com/docs/cloud/settings/telemetry.md): Telemetry endpoints, trace links, the page that draws a conversation as shown, and the SDK clients sending data to a pro

### copilots

- [Assistant Frame API](https://www.assistant-ui.com/docs/copilots/assistant-frame.md): Share model context across iframe boundaries
- [makeAssistantVisible](https://www.assistant-ui.com/docs/copilots/make-assistant-visible.md): Make React components visible and interactive to assistants via higher-order component wrapping.
- [Model Context](https://www.assistant-ui.com/docs/copilots/model-context.md): Configure assistant behavior through system instructions, tools, and context providers.
- [Intelligent Components](https://www.assistant-ui.com/docs/copilots/motivation.md): Add intelligence to React components through readable interfaces and assistant tools.
- [useAssistantInstructions](https://www.assistant-ui.com/docs/copilots/use-assistant-instructions.md): React hook for setting system instructions to guide assistant behavior.

### guides

- [File Attachments](https://www.assistant-ui.com/docs/guides/attachments.md): Let users attach images, PDFs, and other files to AI chat messages in React. Drag-drop, paste, and vision-model support,
- [Message Branching](https://www.assistant-ui.com/docs/guides/branching.md): Edit messages or regenerate AI responses, then switch between alternative replies. Branching navigation built into assis
- [Chain of Thought UI](https://www.assistant-ui.com/docs/guides/chain-of-thought.md): Show AI reasoning steps and tool calls in a collapsible thinking accordion. Build chain-of-thought visualizations in Rea
- [ChatGPT Subscription](https://www.assistant-ui.com/docs/guides/chatgpt-subscription.md): Run your assistant-ui app on your ChatGPT Plus or Pro subscription via Codex OAuth. Local development without an OpenAI 
- [Assistant Context API](https://www.assistant-ui.com/docs/guides/context-api.md): Read and update assistant state to build custom React components in your chat UI — composable context API for thread, me
- [Speech-to-Text Dictation](https://www.assistant-ui.com/docs/guides/dictation.md): Add voice dictation to your AI chat composer with the Web Speech API or a custom adapter. Speech-to-text in React, integ
- [Message Editing](https://www.assistant-ui.com/docs/guides/editing.md): Let users edit their messages and regenerate AI responses with custom editor interfaces. Edit-and-resubmit patterns for 
- [Electron](https://www.assistant-ui.com/docs/guides/electron.md): Run assistant-ui in an Electron renderer with a hosted backend or an AI SDK route in the main process.
- [Headless Composer Input](https://www.assistant-ui.com/docs/guides/headless-composer-input.md): Build a custom composer input while keeping assistant-ui composer state and send gating.
- [Image Generation](https://www.assistant-ui.com/docs/guides/image-generation.md): Generate images in your backend and render them inline in an assistant-ui thread.
- [Guides](https://www.assistant-ui.com/docs/guides.md): Practical recipes for building AI chat features in React with assistant-ui — attachments, branching, multi-agent, voice,
- [Input History](https://www.assistant-ui.com/docs/guides/input-history.md): Terminal-style ArrowUp/ArrowDown recall of previously sent messages in the assistant-ui React composer.
- [LaTeX in Chat Messages](https://www.assistant-ui.com/docs/guides/latex.md): Render LaTeX math expressions in AI chat messages with KaTeX — drop-in equation support for React chat UIs built on assi
- [Mentions in Chat](https://www.assistant-ui.com/docs/guides/mentions.md): Let users @-mention tools or custom items in the AI chat composer to guide the LLM. Mention picker built into assistant-
- [Message Timing & Token Stats](https://www.assistant-ui.com/docs/guides/message-timing.md): Display stream metadata in AI chat — generation duration, tokens per second, and time to first token, rendered via assis
- [OpenAI Realtime](https://www.assistant-ui.com/docs/guides/openai-realtime.md): Connect browser voice conversations directly to OpenAI Realtime with transcripts, interruptions, mute controls, and type
- [Message Part Grouping](https://www.assistant-ui.com/docs/guides/part-grouping.md): Organize message parts into custom groups with flexible grouping functions.
- [Quote Selected Text](https://www.assistant-ui.com/docs/guides/quoting.md): Let users select text from AI messages and quote it back into the composer. Full quoting flow with backend handling and 
- [Resumable Stream Deployment](https://www.assistant-ui.com/docs/guides/resumable-stream-deployment.md): Production hardening for resumable streams. Authorization, serverless lifetimes, TTLs, key isolation, observability, res
- [Custom Resumable Stream Stores](https://www.assistant-ui.com/docs/guides/resumable-stream-stores.md): Implement the ResumableStreamStore interface to back resumable streams with Postgres, Cloudflare Durable Objects, Upstas
- [Resumable Streams](https://www.assistant-ui.com/docs/guides/resumable-streams.md): Persist an in-flight LLM response on the server so the client can reload, lose its connection, or open a new tab and pic
- [Custom Scrollbar](https://www.assistant-ui.com/docs/guides/scrollbar.md): Replace the default scrollbar with a custom Radix UI scroll area.
- [Slash Commands](https://www.assistant-ui.com/docs/guides/slash-commands.md): Trigger predefined actions in your AI chat by typing / — slash command palette with popover, search, and action handlers
- [Text-to-Speech for Chat](https://www.assistant-ui.com/docs/guides/speech.md): Read AI chat messages aloud with the Web Speech API or a custom TTS adapter. Speech synthesis for React chat UIs, integr
- [Reading State Outside the Thread](https://www.assistant-ui.com/docs/guides/state-outside-the-thread.md): Bind a header, a sidebar, a status tray, or code in another React root to thread state without lifting it into a second 
- [Streamdown Markdown Renderer](https://www.assistant-ui.com/docs/guides/streamdown.md): Stream markdown into a React chat UI with syntax highlighting, math, and Mermaid diagrams. Powered by Vercel Streamdown,
- [Suggested Prompts](https://www.assistant-ui.com/docs/guides/suggestions.md): Display suggested starter prompts in your AI chat to onboard users faster. Configurable suggestion components for React,
- [Tauri](https://www.assistant-ui.com/docs/guides/tauri.md): Run assistant-ui in a Tauri webview with a hosted backend or a streaming Rust command.
- [Thread Virtualization](https://www.assistant-ui.com/docs/guides/virtualization.md): Render very long threads with @tanstack/react-virtual, one row per user message, assistant part and turn end, with creat
- [Realtime Voice Chat](https://www.assistant-ui.com/docs/guides/voice.md): Build bidirectional voice conversations with AI in React. Realtime audio streaming, interruption handling, and visual st

### ink

- [Adapters](https://www.assistant-ui.com/docs/ink/adapters.md): Attachment, title generation, and storage adapters for React Ink.
- [Custom Backend](https://www.assistant-ui.com/docs/ink/custom-backend.md): Connect your terminal app to your own backend API.
- [Hooks](https://www.assistant-ui.com/docs/ink/hooks.md): Reactive hooks for accessing runtime state in React Ink.
- [Installation](https://www.assistant-ui.com/docs/ink.md): Get assistant-ui running in a terminal app with your first chat interface.
- [Migration from Web](https://www.assistant-ui.com/docs/ink/migration.md): Migrate an existing @assistant-ui/react app to the terminal with React Ink.
- [Primitives](https://www.assistant-ui.com/docs/ink/primitives.md): Composable terminal components for building chat UIs with Ink.

### integrations

- [Custom attachment uploads](https://www.assistant-ui.com/docs/integrations/attachments/custom-adapter.md): Upload chat attachments to object storage with a presigned-URL AttachmentAdapter.
- [better-auth](https://www.assistant-ui.com/docs/integrations/auth/better-auth.md): TypeScript-first auth with database-owned sessions; gate the chat route and scope threads to the signed-in user.
- [Clerk](https://www.assistant-ui.com/docs/integrations/auth/clerk.md): Gate the chat route and scope thread persistence to the signed-in user with Clerk.
- [Auth.js (next-auth)](https://www.assistant-ui.com/docs/integrations/auth/next-auth.md): Gate the chat route and scope thread persistence to the signed-in user with Auth.js v5.
- [Vercel AI SDK](https://www.assistant-ui.com/docs/integrations/frameworks/ai-sdk.md): Wire the Vercel AI SDK into a React chat UI with assistant-ui, covering useChat, streaming, tools, attachments, multi-st
- [Cloudflare Agents](https://www.assistant-ui.com/docs/integrations/frameworks/cloudflare-agents.md): Connect Cloudflare's stateful agent framework to a React chat UI with assistant-ui through the AI SDK runtime, keeping i
- [Full-stack integration](https://www.assistant-ui.com/docs/integrations/frameworks/mastra/full-stack.md): Run Mastra agents inside your Next.js API routes.
- [Mastra Integration](https://www.assistant-ui.com/docs/integrations/frameworks/mastra/overview.md): Wire the Mastra TypeScript agent framework into a React chat UI with assistant-ui, with full streaming, tool calling, mu
- [Separate server integration](https://www.assistant-ui.com/docs/integrations/frameworks/mastra/separate-server.md): Run Mastra as a standalone server with assistant-ui as a separate frontend.
- [OpenAI Agents SDK](https://www.assistant-ui.com/docs/integrations/frameworks/openai-agents.md): Connect the OpenAI Agents SDK to assistant-ui through the AI SDK runtime.
- [Pydantic AI](https://www.assistant-ui.com/docs/integrations/frameworks/pydantic-ai.md): Connect a Pydantic AI Python agent to assistant-ui through the AG-UI runtime.
- [LLM Gateway Integrations](https://www.assistant-ui.com/docs/integrations/gateways.md): Route AI chat traffic through OpenAI-compatible LLM gateways (OpenRouter, LiteLLM, Portkey, etc.) for cost, fallback, an
- [Integrations](https://www.assistant-ui.com/docs/integrations.md): Adapters for Vercel AI SDK, LangChain, LangGraph, Mastra, plus auth, persistence, observability, and tool services, for 
- [Assistant Cloud](https://www.assistant-ui.com/docs/integrations/observability/assistant-cloud.md): Correlate AI SDK OpenTelemetry spans with Assistant Cloud runs for one view of browser and server activity.
- [Langfuse](https://www.assistant-ui.com/docs/integrations/observability/langfuse.md): Trace AI SDK calls into Langfuse via OpenTelemetry for tracing, evals, and prompt management.
- [LangSmith](https://www.assistant-ui.com/docs/integrations/observability/langsmith.md): Trace AI SDK calls into LangSmith with the wrapAISDK helper.
- [Custom thread persistence](https://www.assistant-ui.com/docs/integrations/persistence/custom-adapter.md): Persist threads and messages to your own database with RemoteThreadListAdapter and ThreadHistoryAdapter.

### migrations

- [Deprecation Policy](https://www.assistant-ui.com/docs/migrations/deprecation-policy.md): Stability guarantees and deprecation timelines for assistant-ui features.
- [Migration Guides](https://www.assistant-ui.com/docs/migrations.md): Upgrade assistant-ui versions and migrate deprecated APIs and integrations.
- [Using old React versions](https://www.assistant-ui.com/docs/migrations/react-compatibility.md): Compatibility notes for React 18 and 19.
- [Migrating to react-langgraph v0.7](https://www.assistant-ui.com/docs/migrations/react-langgraph-v0-7.md): Guide to upgrading to the simplified LangGraph integration API.
- [Migrating Tools to Toolkits](https://www.assistant-ui.com/docs/migrations/toolkit-tools.md): Move makeAssistantTool, useAssistantTool, makeAssistantToolUI, and useAssistantToolUI registrations to the toolkit API.
- [Migration to v0.11](https://www.assistant-ui.com/docs/migrations/v0-11.md): ContentPart renamed to MessagePart for better semantic clarity.
- [Migration to v0.12](https://www.assistant-ui.com/docs/migrations/v0-12.md): Unified state API replaces individual context hooks.
- [Migration to v0.14](https://www.assistant-ui.com/docs/migrations/v0-14.md): Drops APIs deprecated since v0.11/v0.12, and primitives migrate from components prop to children render functions.
- [Migration to v0.15](https://www.assistant-ui.com/docs/migrations/v0-15.md): Drops the v0.12-era legacy runtime hooks, the deprecated tools map, and the "mcp-app" group key. Scope accessors become 

### primitives

- [ActionBar](https://www.assistant-ui.com/docs/primitives/action-bar.md): Build message action buttons with auto-hide, copy state, and intelligent disabling.
- [AssistantModal](https://www.assistant-ui.com/docs/primitives/assistant-modal.md): A floating chat popover with a fixed-position trigger button that opens a chat panel.
- [Attachment](https://www.assistant-ui.com/docs/primitives/attachment.md): File and image attachment rendering for the composer and messages.
- [BranchPicker](https://www.assistant-ui.com/docs/primitives/branch-picker.md): Navigate between message branches, which are alternative responses the user can flip through.
- [ChainOfThought](https://www.assistant-ui.com/docs/primitives/chain-of-thought.md): Collapsible accordion for grouping reasoning steps and tool calls.
- [Composer](https://www.assistant-ui.com/docs/primitives/composer.md): Build custom message input UIs with full control over layout and behavior.
- [Error](https://www.assistant-ui.com/docs/primitives/error.md): Accessible error display for messages with automatic error text extraction.
- [Overview](https://www.assistant-ui.com/docs/primitives.md): Unstyled, accessible building blocks for AI chat interfaces, such as Thread, Composer, and Message.
- [Message](https://www.assistant-ui.com/docs/primitives/message.md): Build custom message rendering with content parts, attachments, and hover state.
- [SelectionToolbar](https://www.assistant-ui.com/docs/primitives/selection-toolbar.md): A floating toolbar that appears when text is selected within a message.
- [Suggestion](https://www.assistant-ui.com/docs/primitives/suggestion.md): Suggested prompts that users can click to quickly send or populate the composer.
- [ThreadList](https://www.assistant-ui.com/docs/primitives/thread-list.md): Multi-thread management for listing, creating, switching, archiving, and deleting conversations.
- [Thread](https://www.assistant-ui.com/docs/primitives/thread.md): Build custom scrollable message containers with auto-scroll, empty states, and message rendering.

### react-native

- [Adapters](https://www.assistant-ui.com/docs/react-native/adapters.md): Persistence and title generation adapters for React Native.
- [Attachments](https://www.assistant-ui.com/docs/react-native/attachments.md): Add image attachments to a React Native composer and send their content to your model.
- [Custom Backend](https://www.assistant-ui.com/docs/react-native/custom-backend.md): Connect your React Native app to your own backend API.
- [Elements](https://www.assistant-ui.com/docs/react-native/elements.md): The React Native elements, what each one installs, and what changes when the same design runs on a phone.
- [Add to an Existing App](https://www.assistant-ui.com/docs/react-native/existing-app.md): Bring the React Native elements into an app that already has its own runtime, its own styles and its own navigation.
- [Generative UI](https://www.assistant-ui.com/docs/react-native/generative-ui.md): Let the model compose cards, tables, charts, and form controls drawn with React Native components.
- [Windowed History](https://www.assistant-ui.com/docs/react-native/history.md): Keep a long thread light by loading older messages above the window as the reader scrolls up.
- [Hooks](https://www.assistant-ui.com/docs/react-native/hooks.md): Reactive hooks for accessing runtime state in React Native.
- [Installation](https://www.assistant-ui.com/docs/react-native.md): Get assistant-ui running in an Expo or React Native app with your first chat screen.
- [Migration from Web](https://www.assistant-ui.com/docs/react-native/migration.md): Migrate an existing @assistant-ui/react app to React Native.
- [Primitives](https://www.assistant-ui.com/docs/react-native/primitives.md): Composable React Native components for building chat UIs.
- [Testing the native kit](https://www.assistant-ui.com/docs/react-native/testing.md): Test copied React Native elements with Vitest, jsdom, and react-native-web.
- [Thread list](https://www.assistant-ui.com/docs/react-native/thread-list.md): Show, select, and create conversations from a React Native thread list.
- [Tool UI and approvals](https://www.assistant-ui.com/docs/react-native/tool-ui.md): Render tool calls, register native tool interfaces, and ask for approval before a tool continues.
- [Static web export](https://www.assistant-ui.com/docs/react-native/web-export.md): Export the Expo example for the web and understand the React Native Web limits that affect chat screens.

### runtimes

- [Client and hooks](https://www.assistant-ui.com/docs/runtimes/a2a/client-and-hooks.md): A2AClient, useA2ARuntime options, hooks, task states, artifacts, errors.
- [A2A Agent Runtime](https://www.assistant-ui.com/docs/runtimes/a2a/overview.md): Connect any A2A v1.0 protocol-compliant agent server to a React chat UI with assistant-ui — full streaming, tool calls, 
- [Quickstart](https://www.assistant-ui.com/docs/runtimes/a2a/quickstart.md): Minimal runtime and Thread setup against an A2A server.
- [Agent state](https://www.assistant-ui.com/docs/runtimes/ag-ui/agent-state.md): Read and optimistically update agent-owned state with useAgUiState and useAgUiSetState over AG-UI.
- [AG-UI Agent Runtime](https://www.assistant-ui.com/docs/runtimes/ag-ui/overview.md): Wire AG-UI (Agent-User Interaction) protocol agents into a React chat UI with assistant-ui — bidirectional events, gener
- [Quickstart](https://www.assistant-ui.com/docs/runtimes/ag-ui/quickstart.md): Minimal HttpAgent + useAgUiRuntime setup against an AG-UI server.
- [Runtime options](https://www.assistant-ui.com/docs/runtimes/ag-ui/runtime-options.md): useAgUiRuntime options, adapters, message conversion, thread list.
- [Vercel AI SDK Runtime](https://www.assistant-ui.com/docs/runtimes/ai-sdk/overview.md): Connect the Vercel AI SDK to a React chat UI via assistant-ui — useChat hooks, custom transports, frontend tools, attach
- [AI SDK v4 (legacy)](https://www.assistant-ui.com/docs/runtimes/ai-sdk/v4-legacy.md): Reference for projects still on AI SDK v4. New projects should use v7.
- [AI SDK v5 (legacy)](https://www.assistant-ui.com/docs/runtimes/ai-sdk/v5-legacy.md): Reference for projects still on AI SDK v5. New projects should use v7.
- [AI SDK v6 (legacy)](https://www.assistant-ui.com/docs/runtimes/ai-sdk/v6-legacy.md): Reference for projects still on AI SDK v6. New projects should use v7.
- [AI SDK v7](https://www.assistant-ui.com/docs/runtimes/ai-sdk/v7.md): Integrate Vercel AI SDK v7 with assistant-ui for streaming chat.
- [Claude Managed Agents](https://www.assistant-ui.com/docs/runtimes/claude-managed-agents.md): Connect Anthropic's Managed Agents sessions to assistant-ui with the external store runtime, folding the session event l
- [Adapters](https://www.assistant-ui.com/docs/runtimes/concepts/adapters.md): Reusable extension points for attachments, speech, feedback, history, and suggestions.
- [Runtime architecture](https://www.assistant-ui.com/docs/runtimes/concepts/architecture.md): How core runtimes, protocol layers, and framework adapters fit together.
- [Stability](https://www.assistant-ui.com/docs/runtimes/concepts/stability.md): What unstable_ means, when APIs become stable, and how to track changes.
- [Threads](https://www.assistant-ui.com/docs/runtimes/concepts/threads.md): Single-thread, cloud, and custom-database thread management.
- [Assistant Transport](https://www.assistant-ui.com/docs/runtimes/custom/assistant-transport.md): Stream agent state to the frontend and handle user commands for custom agents.
- [Data Stream Protocol](https://www.assistant-ui.com/docs/runtimes/custom/data-stream.md): Standard message-streaming protocol on top of LocalRuntime.
- [ExternalStoreRuntime](https://www.assistant-ui.com/docs/runtimes/custom/external-store.md): Bring your own redux, zustand, or state manager.
- [LocalRuntime](https://www.assistant-ui.com/docs/runtimes/custom/local-runtime.md): Quickest path to a working chat. Handles state while you handle the API.
- [Custom Runtime](https://www.assistant-ui.com/docs/runtimes/custom/overview.md): Build a React chat UI for any AI backend with assistant-ui — four runtime patterns covering local state, REST, custom pr
- [Eve Runtime](https://www.assistant-ui.com/docs/runtimes/eve/overview.md): Connect an Eve agent to assistant-ui with useEveAgentRuntime, eve/next, durable sessions, streaming messages, and human-
- [Quickstart](https://www.assistant-ui.com/docs/runtimes/eve/quickstart.md): Template, Eve CLI, and manual setup paths to a working Eve agent chat in assistant-ui.
- [API reference](https://www.assistant-ui.com/docs/runtimes/google-adk/api.md): createAdkStream, server helpers, session adapter, threads, message editing.
- [Hooks](https://www.assistant-ui.com/docs/runtimes/google-adk/hooks.md): Tool confirmations, auth, input requests, artifacts, escalation, metadata, structured events.
- [Google ADK Runtime](https://www.assistant-ui.com/docs/runtimes/google-adk/overview.md): Connect Google's Agent Development Kit (ADK) to a React chat UI with assistant-ui — streaming, tool calls, and multi-age
- [Quickstart](https://www.assistant-ui.com/docs/runtimes/google-adk/quickstart.md): Minimal API route and client setup with createAdkApiRoute.
- [LangChain React Runtime](https://www.assistant-ui.com/docs/runtimes/langchain.md): Use LangChain's useStream hook with a React chat UI through assistant-ui — a lighter LangGraph adapter that delegates st
- [Agent state](https://www.assistant-ui.com/docs/runtimes/langgraph/agent-state.md): Read and optimistically update graph state with useLangGraphState and useLangGraphSetState in LangGraph.
- [LangGraph Generative UI](https://www.assistant-ui.com/docs/runtimes/langgraph/generative-ui.md): Render structured UI components emitted by LangGraph alongside assistant messages.
- [Interrupts and message editing](https://www.assistant-ui.com/docs/runtimes/langgraph/interrupts.md): Interrupt persistence and checkpoint-based message editing.
- [LangGraph UI Runtime](https://www.assistant-ui.com/docs/runtimes/langgraph/overview.md): Build a chat UI for LangGraph agents in React with assistant-ui — streaming, subgraph events, UI messages, interrupts, a
- [Quickstart](https://www.assistant-ui.com/docs/runtimes/langgraph/quickstart.md): From-template and manual setup paths to a working LangGraph chat.
- [Streaming](https://www.assistant-ui.com/docs/runtimes/langgraph/streaming.md): Event handlers, message accumulator, conversion, metadata, and generative UI.
- [Threads](https://www.assistant-ui.com/docs/runtimes/langgraph/threads.md): Basic thread support, AssistantCloud, and custom thread list adapter.
- [Introduction](https://www.assistant-ui.com/docs/runtimes/langgraph/tutorial/introduction.md): Build a stockbroker assistant with LangGraph and assistant-ui.
- [Part 1: Setup frontend](https://www.assistant-ui.com/docs/runtimes/langgraph/tutorial/part-1.md): Create a Next.js project with the LangGraph assistant-ui template.
- [Part 2: Generative UI](https://www.assistant-ui.com/docs/runtimes/langgraph/tutorial/part-2.md): Display stock ticker information with generative UI components.
- [Part 3: Approval UI](https://www.assistant-ui.com/docs/runtimes/langgraph/tutorial/part-3.md): Add human-in-the-loop approval for tool calls.
- [Hooks](https://www.assistant-ui.com/docs/runtimes/opencode/hooks.md): Permissions, questions, session state, runtime extras.
- [OpenCode Runtime](https://www.assistant-ui.com/docs/runtimes/opencode/overview.md): Build a React chat UI for OpenCode coding agents with assistant-ui — streaming, tool calls, file edits, and terminal out
- [Quickstart](https://www.assistant-ui.com/docs/runtimes/opencode/quickstart.md): Minimal useOpenCodeRuntime setup against a local OpenCode server.
- [Picking a runtime](https://www.assistant-ui.com/docs/runtimes/pick-a-runtime.md): Decision guide for choosing the right runtime, by framework or by feature.

### store

- [API Reference](https://www.assistant-ui.com/docs/store/api-reference.md): All exports from @assistant-ui/store.
- [Child Scopes](https://www.assistant-ui.com/docs/store/child-scopes.md): Derive scopes from parent data with Derived, useClientResource, and useClientLookup.
- [Events](https://www.assistant-ui.com/docs/store/events.md): Emit and subscribe to typed events.
- [Meta](https://www.assistant-ui.com/docs/store/meta.md): Track scope origin with source and query.
- [Methods](https://www.assistant-ui.com/docs/store/methods.md): Access scope methods with useAui.
- [Quickstart](https://www.assistant-ui.com/docs/store/quickstart.md): Install Store and connect your first Tap resource to React.
- [Rendering Lists](https://www.assistant-ui.com/docs/store/rendering-lists.md): How to efficiently render lists of items from store state.
- [Scopes](https://www.assistant-ui.com/docs/store/scopes.md): Named, independent units of state in your store.
- [Sibling Scopes](https://www.assistant-ui.com/docs/store/sibling-scopes.md): Scopes that reference each other at the same level.
- [State](https://www.assistant-ui.com/docs/store/state.md): Subscribe to state changes with useAuiState.
- [Why Store](https://www.assistant-ui.com/docs/store/why-store.md): The problem Store solves.

### tap

- [API Reference](https://www.assistant-ui.com/docs/tap/api-reference.md): Public API exported from @assistant-ui/tap.
- [Composition](https://www.assistant-ui.com/docs/tap/composition.md): Combine Resources into reusable state trees.
- [Context](https://www.assistant-ui.com/docs/tap/context.md): Pass values through resource boundaries without prop drilling.
- [How tap differs from React](https://www.assistant-ui.com/docs/tap/differences-from-react.md): The handful of behaviors that aren't quite React.
- [Introduction](https://www.assistant-ui.com/docs/tap.md): State management based on React Hooks.
- [Lifecycle](https://www.assistant-ui.com/docs/tap/lifecycle.md): How resources render, mount, update, and unmount.
- [Motivation](https://www.assistant-ui.com/docs/tap/motivation.md): Composable configuration and state built on React's lifecycle.
- [Outside React](https://www.assistant-ui.com/docs/tap/outside-react.md): Run resources standalone, with no React tree.
- [Quickstart](https://www.assistant-ui.com/docs/tap/quickstart.md): Install tap and build your first Resource.
- [Resources](https://www.assistant-ui.com/docs/tap/resources.md): Package stateful behavior into reusable, configurable values.
- [Trees & Re-renders](https://www.assistant-ui.com/docs/tap/trees-and-rerenders.md): How resource trees re-render and where scheduling boundaries form.

### tools

- [A2UI over AG-UI](https://www.assistant-ui.com/docs/tools/a2ui.md): Render A2UI surfaces from AG-UI activity snapshots or A2A messages, task snapshots, and artifacts as generative UI.
- [Backend Tools](https://www.assistant-ui.com/docs/tools/backend.md): Wire assistant-ui toolkits into your server with the AI SDK using AISDKToolkit and frontendTools, mix client and server 
- [Defining Tools](https://www.assistant-ui.com/docs/tools/defining-tools.md): Define tools for your AI chat with assistant-ui toolkits and the "use generative" directive, covering frontend, backend,
- [Dynamic Tools](https://www.assistant-ui.com/docs/tools/dynamic-tools.md): Tools whose executor closes over React state. Declare the contract with stubTool() in a "use generative" file and supply
- [Choosing an answer format](https://www.assistant-ui.com/docs/tools/generative-ui-formats.md): When the model should answer in Markdown, with spec mode or the present tool, or with a widget in a sandboxed frame, and
- [Generative UI primitive](https://www.assistant-ui.com/docs/tools/generative-ui-primitive.md): Render agent-described React UI from a JSON spec with a consumer-provided component allowlist, using the MessagePrimitiv
- [Generative UI on Slack](https://www.assistant-ui.com/docs/tools/generative-ui-slack.md): Convert a generative UI tree into Slack Block Kit, post it, and decode the block_actions payload Slack sends back into y
- [Moving to spec mode](https://www.assistant-ui.com/docs/tools/generative-ui-spec-migration.md): Map a present tree, its actions, and its setup onto generative-frame spec mode, and keep present for what spec mode does
- [Generative UI on Microsoft Teams](https://www.assistant-ui.com/docs/tools/generative-ui-teams.md): Convert a generative UI tree into an Adaptive Card, send it from a bot, and decode the Action.Submit payload Teams sends
- [Generative UI](https://www.assistant-ui.com/docs/tools/generative-ui.md): Let the model compose an interface at runtime from a component vocabulary you ship, using the present tool from @assista
- [Tools](https://www.assistant-ui.com/docs/tools.md): Give the model callable capabilities with assistant-ui toolkits. Define frontend, backend, human, and provider tools, re
- [Interactable Tool UIs](https://www.assistant-ui.com/docs/tools/interactables.md): Build stateful components and tool UIs that both the user and the model can read and edit. Render them beside the thread
- [MCP Apps](https://www.assistant-ui.com/docs/tools/mcp-apps.md): Render MCP App UI resources inline in chat. Native renderer for the Model Context Protocol Apps spec — sandboxed iframes
- [Model Context Protocol (MCP)](https://www.assistant-ui.com/docs/tools/mcp.md): Connect MCP servers as a tool catalog in your assistant-ui app.
- [Multi-Agent Chat UI](https://www.assistant-ui.com/docs/tools/multi-agent.md): Render sub-agent conversations and handoffs inside tool calls. Build supervisor and multi-agent patterns in a React chat
- [OpenUI](https://www.assistant-ui.com/docs/tools/openui.md): Render streaming OpenUI Lang interfaces in an assistant-ui conversation with @openuidev/assistant-ui, the integration pa
- [Tool UI](https://www.assistant-ui.com/docs/tools/tool-ui.md): Render AI tool calls as custom React components that show loading, result, and interactive states for each tool invocati
- [User-managed MCP servers](https://www.assistant-ui.com/docs/tools/user-managed-mcp.md): Let end users add and authenticate MCP servers from the browser with @assistant-ui/react-mcp.
- [WebMCP provider](https://www.assistant-ui.com/docs/tools/webmcp.md): Expose your app's frontend tools to a WebMCP-capable browser with unstable_useWebMcpProvider.

### utilities

- [heat-graph](https://www.assistant-ui.com/docs/utilities/heat-graph.md): Headless, composable activity heatmap components for React.
- [react-o11y](https://www.assistant-ui.com/docs/utilities/react-o11y.md): Headless primitives for visualizing observability spans as collapsible trace trees, waterfalls, and agent and LLM call t
- [tw-shimmer](https://www.assistant-ui.com/docs/utilities/tw-shimmer.md): Tailwind CSS v4 plugin for shimmer effects.

### vue

- [Architecture](https://www.assistant-ui.com/docs/vue/architecture.md): How AuiProvider, the assistant client, its scopes, and the Vue composables fit together.
- [Attachments](https://www.assistant-ui.com/docs/vue/attachments.md): Let users attach files to a message and control which files the runtime accepts and how it sends them.
- [Branching](https://www.assistant-ui.com/docs/vue/branching.md): Keep every edited and regenerated version of a message and let users move between them.
- [Editing](https://www.assistant-ui.com/docs/vue/editing.md): Let users edit a sent message and regenerate a response.
- [Generative UI](https://www.assistant-ui.com/docs/vue/generative-ui.md): Let the model answer with a spec of cards, tables, and form controls that renders in the thread with Vue components.
- [Markdown](https://www.assistant-ui.com/docs/vue/markdown.md): Render assistant text as Markdown with the registry component or a renderer of your own.
- [Model context](https://www.assistant-ui.com/docs/vue/model-context.md): Send system instructions, tools, and call settings with each request, and change them while the app runs.
- [Reasoning](https://www.assistant-ui.com/docs/vue/reasoning.md): Stream the model's reasoning into a collapsible block next to the answer.
- [State and actions](https://www.assistant-ui.com/docs/vue/state.md): Read assistant state with useAuiState, call actions through useAui, and react to events with useAuiEvent.
- [Suggestions](https://www.assistant-ui.com/docs/vue/suggestions.md): Offer starter prompts and follow-up suggestions, from static config or generated by the runtime.
- [Thread list](https://www.assistant-ui.com/docs/vue/thread-list.md): Let users switch between conversations, start new ones, and archive, rename, or delete old ones.
- [Tool UI](https://www.assistant-ui.com/docs/vue/tool-ui.md): Render tool calls and data parts with Vue components, and answer tool calls that wait for the user.
- [Defining tools](https://www.assistant-ui.com/docs/vue/tools.md): Give the model tools that run on your server or in the browser, and tools that wait for the user.
- [AI SDK](https://www.assistant-ui.com/docs/vue/ai-sdk.md): Stream from an AI SDK route with AISDKChat or AISDKThreads, and configure the transport, callbacks, and adapters.
- [Cloud persistence](https://www.assistant-ui.com/docs/vue/cloud.md): Store threads, messages, titles, and attachments in assistant-cloud with AISDKThreads.
- [External store](https://www.assistant-ui.com/docs/vue/external-store.md): Drive the Vue primitives from messages your own code owns, through RuntimeAdapter.
- [Runtimes](https://www.assistant-ui.com/docs/vue/runtimes.md): Choose how the Vue primitives get their messages: the AI SDK, assistant-cloud, or your own store.
- [Assistant client](https://www.assistant-ui.com/docs/vue/api-reference/client.md): AuiConfig, Derived, createAssistantClient, and the constructors that fill each config key.
- [Components and composables](https://www.assistant-ui.com/docs/vue/api-reference/composables.md): AuiProvider, AuiIf, the scope providers, and the useAui, useAuiState, useAuiEvent, and useScrollLock composables.
- [Events](https://www.assistant-ui.com/docs/vue/api-reference/events.md): Event names, payloads, and scope selectors for useAuiEvent.
- [State and methods](https://www.assistant-ui.com/docs/vue/api-reference/state.md): Every scope's state fields and methods, as read with useAuiState and called through useAui.
- [Styled components](https://www.assistant-ui.com/docs/vue/components.md): Install the registry thread and thread list, then restyle or extend the copied components.
- [Introduction](https://www.assistant-ui.com/docs/vue.md): Build streaming AI chat interfaces with Vue, Nuxt, and the AI SDK.
- [Installation](https://www.assistant-ui.com/docs/vue/installation.md): Add @assistant-ui/vue to an existing Vue or Nuxt app, with or without the AI SDK.
- [Action bar](https://www.assistant-ui.com/docs/vue/primitives/action-bar.md): Add copy, edit, and regenerate buttons to a message.
- [Attachment](https://www.assistant-ui.com/docs/vue/primitives/attachment.md): Render one composer or message attachment with its name, type label, and remove button.
- [Branch picker](https://www.assistant-ui.com/docs/vue/primitives/branch-picker.md): Show which version of a message is displayed and switch between versions.
- [Chain of thought](https://www.assistant-ui.com/docs/vue/primitives/chain-of-thought.md): Group reasoning and tool-call parts into one collapsible block.
- [Composer](https://www.assistant-ui.com/docs/vue/primitives/composer.md): Build the message input, the send and stop buttons, attachment controls, and in-place message editing.
- [Error](https://www.assistant-ui.com/docs/vue/primitives/error.md): Show the error that ended an assistant response.
- [Using primitives](https://www.assistant-ui.com/docs/vue/primitives.md): Unstyled Vue components that bind native elements to assistant state, and the conventions they share.
- [Message](https://www.assistant-ui.com/docs/vue/primitives/message.md): Render one message with its parts, registered tool and data UI, attachments, and hover state.
- [Suggestion](https://www.assistant-ui.com/docs/vue/primitives/suggestion.md): Render suggested prompts that fill the composer or send right away.
- [Thread list](https://www.assistant-ui.com/docs/vue/primitives/thread-list.md): Render the conversation list, the new thread button, and per-thread actions.
- [Thread](https://www.assistant-ui.com/docs/vue/primitives/thread.md): Build the scrollable message list with auto-scroll, a scroll-to-bottom button, a pinned footer, and an empty state.
- [Quickstart](https://www.assistant-ui.com/docs/vue/quickstart.md): Build a streaming Nuxt chat with Vue, assistant-ui, and the AI SDK.
- [Server rendering](https://www.assistant-ui.com/docs/vue/ssr.md): Keep Vue assistant state in a client-only Nuxt component and stream from Nitro.

### examples

- [AI SDK Chat Persistence](https://www.assistant-ui.com/examples/ai-sdk.md): Vercel AI SDK chat with thread persistence. Open-source React example combining the AI SDK and assistant-ui for streamin
- [Claude Artifacts Example](https://www.assistant-ui.com/examples/artifacts.md): Open-source Claude Artifacts implementation in React. Generate websites and components in a side panel from chat message
- [ChatGPT Clone Example](https://www.assistant-ui.com/examples/chatgpt.md): Open-source ChatGPT clone built in React with assistant-ui. Centered welcome composer, high-contrast user bubbles, toolt
- [Claude Clone](https://www.assistant-ui.com/examples/claude.md): Open-source Claude clone in React. Warm cream theme, serif typography, hover-only action bars, and a flat bordered compo
- [Expo React Native AI Chat](https://www.assistant-ui.com/examples/expo.md): Native iOS and Android AI chat app with Expo. Drawer navigation, thread persistence, and the assistant-ui React Native c
- [Form-Filling AI Copilot](https://www.assistant-ui.com/examples/form-demo.md): Open-source AI copilot that fills forms for users. Sidebar UI, field-aware tool calls, and a working React example built
- [Gemini Clone](https://www.assistant-ui.com/examples/gemini.md): Open-source Gemini clone in React with a centered greeting over an ambient glow, a single-row pill composer, avatar-free
- [Generative UI Example](https://www.assistant-ui.com/examples/generative-ui.md): Open-source React example where the model answers each question in Markdown, with a spec built from the app's components
- [Grok Clone](https://www.assistant-ui.com/examples/grok.md): Open-source Grok clone in React. Pill composer with paperclip, animated Mic and Send swap, functional model picker dropd
- [Examples](https://www.assistant-ui.com/examples.md): Production-ready examples of AI chat in React. ChatGPT clones, copilots, generative UI, artifacts, multimodal, and more,
- [Mem0 Memory Chat](https://www.assistant-ui.com/examples/mem0.md): AI chat with persistent memory powered by Mem0. It remembers user preferences, facts, and history across sessions. Open-
- [Floating Modal Chat](https://www.assistant-ui.com/examples/modal.md): Embeddable AI assistant in a floating button modal. Drop it into any React app for in-product copilots or support chat, 
- [Perplexity Clone](https://www.assistant-ui.com/examples/perplexity.md): Open-source Perplexity-style chat in React. Theme-aware composer, functional Search and Model dropdowns, four-state prim
- [LangGraph Stockbroker Demo](https://www.assistant-ui.com/examples/stockbroker.md): Human-in-the-loop AI stockbroker built on LangGraph and assistant-ui. Interrupt handling, tool approval, and an interact

### design

- [Accordion](https://www.assistant-ui.com/design/components/accordion.md): Stacked headings that reveal or hide content sections.
- [Avatar](https://www.assistant-ui.com/design/components/avatar.md): People as initials or images, alone or stacked in groups.
- [Badge](https://www.assistant-ui.com/design/components/badge.md): A small label component for displaying status, categories, or metadata.
- [Breadcrumb](https://www.assistant-ui.com/design/components/breadcrumb.md): Where you are, as a path you can walk back up.
- [Button](https://www.assistant-ui.com/design/components/button.md): The pressable object: thin, flat, and 8px round, in every weight a surface needs.
- [Callout](https://www.assistant-ui.com/design/components/callout.md): A quiet aside on a colored rule, for notes, warnings, and errors in prose.
- [Code Block](https://www.assistant-ui.com/design/components/code-block.md): Highlighted code on a field panel, with a title bar and a copy button.
- [Collapsible](https://www.assistant-ui.com/design/components/collapsible.md): One region that folds open and closed in place.
- [Combobox](https://www.assistant-ui.com/design/components/combobox.md): Type to filter, then pick: a select with a search field.
- [Command Tabs](https://www.assistant-ui.com/design/components/command-tabs.md): One command in several dialects, with synced tabs and a copy button.
- [Definition List](https://www.assistant-ui.com/design/components/definition-list.md): A spec sheet of terms and details, one hairline per row.
- [Dialog](https://www.assistant-ui.com/design/components/dialog.md): A modal sheet for decisions that need the page to wait.
- [Diff Viewer](https://www.assistant-ui.com/design/components/diff-viewer.md): Code changes as unified or split diffs, line by line, on the kit's field panel.
- [Dot Matrix](https://www.assistant-ui.com/design/components/dot-matrix.md): Tiny 5x5 dot-matrix indicator with 20 state-specific blink patterns.
- [Dropdown Menu](https://www.assistant-ui.com/design/components/dropdown-menu.md): Actions behind a trigger: items, shortcuts, and separators on a quiet surface.
- [Input](https://www.assistant-ui.com/design/components/input.md): A single-line field with a label, drawn with a hairline.
- [Kbd](https://www.assistant-ui.com/design/components/kbd.md): Keyboard keys as printed caps, for shortcuts and hints.
- [Number Roll](https://www.assistant-ui.com/design/components/number-roll.md): Animated number that rolls digits odometer-style when the value changes.
- [Popover](https://www.assistant-ui.com/design/components/popover.md): Anchored detail on demand, dismissed by clicking away.
- [Scrollbar](https://www.assistant-ui.com/design/components/scrollbar.md): A quiet overlay scrollbar for panes that keep their own scroll.
- [Select](https://www.assistant-ui.com/design/components/select.md): A dropdown select component with composable sub-components.
- [Separator](https://www.assistant-ui.com/design/components/separator.md): A hairline between things, horizontal or vertical.
- [Sheet](https://www.assistant-ui.com/design/components/sheet.md): A panel that slides in from an edge for secondary flows.
- [Skeleton](https://www.assistant-ui.com/design/components/skeleton.md): The loading shape of content that has a known layout.
- [Steps](https://www.assistant-ui.com/design/components/steps.md): A numbered rail for processes that happen in order.
- [Switch](https://www.assistant-ui.com/design/components/switch.md): A binary control in two sizes; the only capsule in the kit.
- [Table](https://www.assistant-ui.com/design/components/table.md): A print-register data table with hairline rules and mono headers.
- [Tabs](https://www.assistant-ui.com/design/components/tabs.md): A tabs component for organizing content into switchable panels.
- [Toast](https://www.assistant-ui.com/design/components/toast.md): A passing message for outcomes: success, error, and undo.
- [Tooltip](https://www.assistant-ui.com/design/components/tooltip.md): A short label that appears on hover and stays out of the way.

### elements

- [Activity graph](https://www.assistant-ui.com/elements/activity-graph.md): A half-year of runs as a calendar of cells, dense where the work was.
- [Agent card](https://www.assistant-ui.com/elements/agent-card.md): Who you are about to talk to: its skills, its model, and the endpoint behind it.
- [Handoff](https://www.assistant-ui.com/elements/agent-handoff.md): Control passing between agents, with the reason and what came along.
- [Agent plan](https://www.assistant-ui.com/elements/agent-plan.md): A checklist the agent works through, with progress you can glance.
- [Agent status](https://www.assistant-ui.com/elements/agent-status.md): One pill that always answers: what is it doing, and for how long.
- [Approval card](https://www.assistant-ui.com/elements/approval-card.md): Human in the loop: the agent asks before it takes a consequential action.
- [Artifact card](https://www.assistant-ui.com/elements/artifact-card.md): A generated document as a tangible object, written live and versioned.
- [Assistant modal](https://www.assistant-ui.com/elements/assistant-modal.md): A floating chat bubble for support widgets, help desks, and embedded assistants, with a thread list and a resizable wind
- [Assistant sidebar](https://www.assistant-ui.com/elements/assistant-sidebar.md): A resizable side panel for copilot experiences and contextual assistance.
- [Attachment](https://www.assistant-ui.com/elements/attachment.md): Runtime attachments for the composer and messages, with previews, progress, and removal.
- [Background runs](https://www.assistant-ui.com/elements/background-inbox.md): Work still going somewhere else, and the results waiting to be collected.
- [Canvas](https://www.assistant-ui.com/elements/canvas-split.md): The thread steps aside and the document takes the room, still being written as you read.
- [Chart](https://www.assistant-ui.com/elements/chart.md): Area, line, and bars, with points landing one at a time as the series streams in.
- [Chat panel](https://www.assistant-ui.com/elements/chat-panel.md): The whole family working together: a message, a pause, a streamed reply.
- [Checkpoints](https://www.assistant-ui.com/elements/checkpoint-history.md): Points you can fall back to, with what each one would give back.
- [Code diff](https://www.assistant-ui.com/elements/code-diff.md): A unified diff with tinted additions and removals, sized for chat.
- [Code runner](https://www.assistant-ui.com/elements/code-runner.md): A snippet with a run button, and the output it produced attached below it.
- [Command palette](https://www.assistant-ui.com/elements/command-palette.md): Everything the app can do, one keystroke away and grouped by where it acts.
- [Comparison](https://www.assistant-ui.com/elements/comparison-card.md): Two options weighed side by side, with the pick named and argued.
- [Attachments](https://www.assistant-ui.com/elements/composer-attachments.md): Files stage inside the composer with per-file progress before the message sends.
- [Context](https://www.assistant-ui.com/elements/composer-context.md): A token ring in the rail fills as the conversation grows, warning near the limit.
- [Mentions](https://www.assistant-ui.com/elements/composer-mentions.md): Type @ to pull people and agents into the conversation, filtered as you go.
- [Models](https://www.assistant-ui.com/elements/composer-model-picker.md): The model lives in the composer rail, one tap away with context at a glance.
- [Slash commands](https://www.assistant-ui.com/elements/composer-slash-commands.md): Type a slash and the command menu floats above the input, filtering as you continue.
- [Composer trigger popover](https://www.assistant-ui.com/elements/composer-trigger-popover.md): A character-triggered picker for mentions, slash commands, and nested composer actions.
- [Dictation](https://www.assistant-ui.com/elements/composer-voice.md): The mic morphs the input into a live waveform, then lands the transcript as text.
- [Composer](https://www.assistant-ui.com/elements/composer.md): The unified input: attachments, commands, mentions, models, voice, and context in one surface.
- [Computer use](https://www.assistant-ui.com/elements/computer-use.md): The screen the agent is driving, with a cursor trail and what it is doing right now.
- [Confidence](https://www.assistant-ui.com/elements/confidence-marker.md): Which claims came from a source, which were inferred, and which are guesses.
- [Connection state](https://www.assistant-ui.com/elements/connection-state.md): The socket drops, the run keeps going on the server, and the stream is picked back up.
- [Context breakdown](https://www.assistant-ui.com/elements/context-breakdown.md): Where the window actually went: prompt, tools, files, conversation, and what's left.
- [Context display](https://www.assistant-ui.com/elements/context-display.md): Model context usage as a ring, bar, or text value with a detailed hover view.
- [Search in conversation](https://www.assistant-ui.com/elements/conversation-search.md): Find inside a long thread, with every hit marked down the scrollbar.
- [Cost meter](https://www.assistant-ui.com/elements/cost-meter.md): What the run spent, split by model, against the session total.
- [Data table](https://www.assistant-ui.com/elements/data-table.md): A compact, sortable table for structured tool results and application data.
- [Timestamps](https://www.assistant-ui.com/elements/day-separator.md): Chronology in a long thread: days marked, times on hover.
- [Diagram](https://www.assistant-ui.com/elements/diagram.md): A drawn answer with zoom, reset, and a full-bleed view; you hand it the rendered graphic.
- [Directive text](https://www.assistant-ui.com/elements/directive-text.md): A message renderer that turns mention directives into inline, runtime-aware chips.
- [Document reference](https://www.assistant-ui.com/elements/document-reference.md): A document the answer leans on, with the quoted passage and the page to jump to.
- [Draft restore](https://www.assistant-ui.com/elements/draft-restore.md): Come back to a thread and the sentence you never sent is still waiting.
- [Edit a sent message](https://www.assistant-ui.com/elements/edit-message.md): Rewrite a turn in place, told up front how many replies the edit throws away.
- [Elicitation form](https://www.assistant-ui.com/elements/elicitation-form.md): A server pausing mid-tool-call to ask you for the fields it still needs.
- [Empty state](https://www.assistant-ui.com/elements/empty-state.md): The first screen: a greeting, three ways in, and the composer front and center.
- [Error state](https://www.assistant-ui.com/elements/error-state.md): A quiet failure banner with a retry path, not a modal in your face.
- [Feedback dialog](https://www.assistant-ui.com/elements/feedback-dialog.md): A thumbs-down that asks why, so the signal arrives with a reason attached.
- [File tree](https://www.assistant-ui.com/elements/file-tree.md): Everything a run touched, as a tree, with the churn spelled out per file.
- [File](https://www.assistant-ui.com/elements/file.md): File message parts with type-aware icons, inline media playback, filename, size, and download actions.
- [Flow graph](https://www.assistant-ui.com/elements/flow-graph.md): Work as a graph rather than a list: branches that fan out and rejoin.
- [Follow-up suggestions](https://www.assistant-ui.com/elements/follow-up-suggestions.md): Prompt chips populated from the runtime's generated follow-up suggestions.
- [Generative UI](https://www.assistant-ui.com/elements/generative-ui.md): A styled component library for rendering structured generative UI output.
- [Geo map](https://www.assistant-ui.com/elements/geo-map.md): Real coordinates become a tiled map, route lines, and an accessible list of the places returned by a tool.
- [Guardrail notice](https://www.assistant-ui.com/elements/guardrail-notice.md): A refusal in its own shape, with the nearest thing it can do instead.
- [Heat graph](https://www.assistant-ui.com/elements/heat-graph.md): An activity heat map with month labels, weekday labels, legend, and tooltip.
- [Image gallery](https://www.assistant-ui.com/elements/image-gallery.md): A compact image grid that opens every result in one shared lightbox.
- [Image generation](https://www.assistant-ui.com/elements/image-generation.md): A dot grid holds the frame while the image resolves out of a blur.
- [Image](https://www.assistant-ui.com/elements/image.md): Image message parts with preview, loading states, actions, and a fullscreen view.
- [Inline citation](https://www.assistant-ui.com/elements/inline-citation.md): A compact, numbered source marker you place beside any claim in prose.
- [Job progress](https://www.assistant-ui.com/elements/job-progress.md): Weighted job stages that announce live progress and settle into a receipt.
- [Launcher](https://www.assistant-ui.com/elements/launcher-bubble.md): The floating entry point, and the panel it opens into.
- [Link preview](https://www.assistant-ui.com/elements/link-preview.md): A compact card that gives a returned URL enough context to decide whether to open it.
- [Loader](https://www.assistant-ui.com/elements/loading-state.md): A pixel matrix that keeps time while the model has nothing to show yet.
- [Model logos](https://www.assistant-ui.com/elements/logos.md): Inline SVG marks for OpenAI, Anthropic, and Google model providers.
- [Map](https://www.assistant-ui.com/elements/map-answer.md): A location answer: pins, a route between them, and the list they came from.
- [Markdown text](https://www.assistant-ui.com/elements/markdown-text.md): Assistant markdown with headings, lists, links, tables, and code blocks.
- [Math](https://www.assistant-ui.com/elements/math-block.md): Rendered expressions with the working shown, one step at a time.
- [MCP config dialog](https://www.assistant-ui.com/elements/mcp-config.md): A dialog for connectors and custom MCP servers, including authentication and connection state.
- [Server panel](https://www.assistant-ui.com/elements/mcp-server-panel.md): Which servers are connected, what each one brought, and which is still waiting on you.
- [Media player](https://www.assistant-ui.com/elements/media-player.md): Compact inline audio and video players for files and standalone media URLs.
- [Memory](https://www.assistant-ui.com/elements/memory-chips.md): What it now remembers about you, written during the turn and removable.
- [Mermaid diagram](https://www.assistant-ui.com/elements/mermaid-diagram.md): Mermaid diagrams rendered inside messages, including partial streaming input.
- [Message actions](https://www.assistant-ui.com/elements/message-actions.md): Copy, rate, and regenerate. Each action confirms itself with a small state change.
- [Message branches](https://www.assistant-ui.com/elements/message-branches.md): Navigate between regenerated versions of the same answer without losing your place.
- [Message pair](https://www.assistant-ui.com/elements/message-pair.md): A user bubble and a streaming assistant reply, with actions that appear on hover.
- [Message queue](https://www.assistant-ui.com/elements/message-queue.md): Turns you typed while a run was in flight, stacked and cancelable until it finishes.
- [Message timing](https://www.assistant-ui.com/elements/message-timing.md): Streaming statistics for the current message, including first token, total time, and speed.
- [Mobile composer](https://www.assistant-ui.com/elements/mobile-composer.md): The bottom sheet: keyboard-aware, quick actions above, thumb-sized targets.
- [Model selector](https://www.assistant-ui.com/elements/model-selector.md): A searchable runtime model picker with grouped providers and reasoning effort controls.
- [Number ticker](https://www.assistant-ui.com/elements/number-ticker.md): Digits that roll into place as a count updates in real time.
- [Onboarding](https://www.assistant-ui.com/elements/onboarding.md): First run: three moves that teach what this assistant is actually for.
- [Option list](https://www.assistant-ui.com/elements/option-list.md): The agent asks a question with a few answers; the pick returns to it and stays as a receipt.
- [Orb](https://www.assistant-ui.com/elements/orb.md): The realtime voice orb, with connection, mute, and speaking state controls.
- [Permission grant](https://www.assistant-ui.com/elements/permission-grant.md): Granting a capability rather than approving one action, with the reach spelled out.
- [Prompt library](https://www.assistant-ui.com/elements/prompt-library.md): Prompts you saved, searchable, with their variables shown before you insert one.
- [Question flow](https://www.assistant-ui.com/elements/question-flow.md): Ask a few short questions in sequence and return every answer as one result.
- [Quota banner](https://www.assistant-ui.com/elements/quota-banner.md): How much is left, when it comes back, and the way to get more.
- [Quote](https://www.assistant-ui.com/elements/quote.md): Select message text, quote it from a floating toolbar, and carry it into the composer.
- [Read aloud](https://www.assistant-ui.com/elements/read-aloud.md): An answer played back, the spoken word lit as it goes, speed under your thumb.
- [Reasoning effort](https://www.assistant-ui.com/elements/reasoning-effort.md): How hard to think, and how much of that budget the run actually spent.
- [Reasoning](https://www.assistant-ui.com/elements/reasoning.md): A collapsible renderer for assistant reasoning that follows the active message part.
- [Recommendation card](https://www.assistant-ui.com/elements/recommendation-card.md): The agent proposes a change with its confidence, and waits for a yes.
- [Regenerate with](https://www.assistant-ui.com/elements/regenerate-menu.md): Fork the same turn to a different model instead of rolling the same dice.
- [Research report](https://www.assistant-ui.com/elements/research-report.md): An outline that fills in section by section, each carrying the sources behind it.
- [Retrieval chunks](https://www.assistant-ui.com/elements/retrieval-chunks.md): The passages a retrieval answer stands on, scored, before the answer itself arrives.
- [Reviewable diff](https://www.assistant-ui.com/elements/reviewable-diff.md): The same diff, but each hunk is a decision: keep it, discard it, apply what survived.
- [Run activity](https://www.assistant-ui.com/elements/run-activity.md): A compact live summary, ordered commentary and tools, and a final answer that stays visible.
- [Schedule](https://www.assistant-ui.com/elements/schedule-card.md): A run that repeats on its own, with its cadence and how it has been doing.
- [Score breakdown](https://www.assistant-ui.com/elements/score-breakdown.md): A verdict with its arithmetic shown: criteria, weights, and what pulled it down.
- [Scroll anchor](https://www.assistant-ui.com/elements/scroll-anchor.md): Streaming never steals your scroll position; a pill offers the way back down.
- [Settings](https://www.assistant-ui.com/elements/settings-panel.md): Model, system prompt, temperature, and what the assistant is allowed to do.
- [Shared conversation](https://www.assistant-ui.com/elements/shared-conversation.md): A read-only transcript someone sent you, with a way to pick it up yourself.
- [Shiki highlighter](https://www.assistant-ui.com/elements/shiki-highlighter.md): Shiki code highlighting that defers tokenization until a message part settles.
- [Sources](https://www.assistant-ui.com/elements/sources.md): Safe runtime citation chips and static source cards with linked reference details.
- [Speaker identity](https://www.assistant-ui.com/elements/speaker-identity.md): Who is talking, once a thread holds more than a user and one model.
- [Spec sheet](https://www.assistant-ui.com/elements/spec-sheet.md): The most common structured answer after a table: one object, labeled.
- [Stopped run](https://www.assistant-ui.com/elements/stopped-run.md): You pressed stop. The half-written answer stays, and continuing is one tap away.
- [Streaming text](https://www.assistant-ui.com/elements/streaming-text.md): Tokens arrive softly: the newest words land in blue and settle into ink.
- [Subagent list](https://www.assistant-ui.com/elements/subagent-list.md): Parallel workers with their own progress, models, and completions.
- [Syntax highlighter](https://www.assistant-ui.com/elements/syntax-highlighter.md): Prism-based code highlighting for assistant markdown code blocks.
- [Task card](https://www.assistant-ui.com/elements/task-card.md): A delegated task with its state, timing, result, and transcript in one card.
- [Terminal block](https://www.assistant-ui.com/elements/terminal-block.md): Command output that streams line by line and ends with an exit status.
- [Thinking indicator](https://www.assistant-ui.com/elements/thinking-indicator.md): A live status line that names what the agent is doing right now, with elapsed time.
- [Thread list sidebar](https://www.assistant-ui.com/elements/thread-list-sidebar.md): A complete sidebar shell that places the runtime thread list beside the active conversation.
- [Thread list](https://www.assistant-ui.com/elements/thread-list.md): Runtime-backed conversation switching with search, active selection, and thread actions.
- [Thread search](https://www.assistant-ui.com/elements/thread-search.md): History you can actually get back into: pinned first, then grouped by when.
- [Thread](https://www.assistant-ui.com/elements/thread.md): A complete chat container with messages, composer, auto-scroll, and accessibility built in.
- [Timeline](https://www.assistant-ui.com/elements/timeline.md): Events on a time axis, with what already happened and what is still coming.
- [Todo list](https://www.assistant-ui.com/elements/todo-list.md): The agent's own working list, rewritten mid-run as it discovers what else is needed.
- [Tool call](https://www.assistant-ui.com/elements/tool-call.md): One tool invocation with its request and result tucked behind a disclosure.
- [Tool failure](https://www.assistant-ui.com/elements/tool-error.md): One call failed. The error, the attempt count, and a retry that doesn't restart the turn.
- [Tool fallback](https://www.assistant-ui.com/elements/tool-fallback.md): The default runtime renderer for tool calls that do not have dedicated UI.
- [Tool group](https://www.assistant-ui.com/elements/tool-group.md): A collapsible runtime wrapper around consecutive tool calls in one assistant turn.
- [Tool timeline](https://www.assistant-ui.com/elements/tool-timeline.md): A whole working session summarized as verbs, targets, and file stats.
- [Tooltip icon button](https://www.assistant-ui.com/elements/tooltip-icon-button.md): An accessible icon button with a tooltip label and shared interaction states.
- [Trace waterfall](https://www.assistant-ui.com/elements/trace-waterfall.md): Every span in a run on one time axis, nested, so you can see where it actually went.
- [Typing indicator](https://www.assistant-ui.com/elements/typing-indicator.md): The classic three dots, tuned to read as presence rather than noise.
- [Voice conversation](https://www.assistant-ui.com/elements/voice-conversation.md): A live call: the orb tracks your voice, the caption names the turn, the transcript follows.
- [Web preview](https://www.assistant-ui.com/elements/web-preview.md): Chrome for a sandboxed preview: a URL bar, reload, and open-in-new around a frame you isolate.
- [Web search](https://www.assistant-ui.com/elements/web-search.md): A search query and its results landing one by one as the agent reads.

### Safe Content Frame

- [API reference](https://www.assistant-ui.com/safe-content-frame/docs/api-reference.md): The SafeContentFrame class, its options, the render methods, and the rendered frame.
- [Usage with Generative UI](https://www.assistant-ui.com/safe-content-frame/docs/generative-ui.md): Render model-generated HTML from a React component, and exchange messages with it.
- [How Safe Content Frame works](https://www.assistant-ui.com/safe-content-frame/docs/how-it-works.md): How safe-content-frame gives every render its own domain, and what that changes compared with an iframe sandbox.
- [Introduction](https://www.assistant-ui.com/safe-content-frame/docs.md): Sandboxes for HTML. Render untrusted content in an iframe that runs on its own domain.
- [Usage with MCP Apps](https://www.assistant-ui.com/safe-content-frame/docs/mcp-apps.md): How the assistant-ui MCP Apps renderer mounts each widget in a Safe Content Frame.

### Generative Frame

- [API reference](https://www.assistant-ui.com/generative-frame/docs/api-reference.md): Signatures and options for every generative-frame entry point.
- [With assistant-ui](https://www.assistant-ui.com/generative-frame/docs/assistant-ui.md): Render widget tool calls in an assistant-ui thread as their arguments stream, and report render errors back to the model
- [Delegate to a sub-agent](https://www.assistant-ui.com/generative-frame/docs/delegation.md): Keep widget code out of the main agent's context by giving it one tool that a second model call fills in, checked with t
- [Introduction](https://www.assistant-ui.com/generative-frame/docs.md): Render model-generated HTML and SVG widgets in a sandboxed frame while they stream.
- [Quickstart](https://www.assistant-ui.com/generative-frame/docs/quickstart.md): Stream a model-generated widget into a page with plain TypeScript and no framework.
- [React](https://www.assistant-ui.com/generative-frame/docs/react.md): Render streaming widgets from React with the Widget component, useWidget, and useThemeTokens.
- [Repair loop and screenshots](https://www.assistant-ui.com/generative-frame/docs/repair.md): Render widget code offscreen, turn errors into model-readable feedback, and capture screenshots inside the frame.
- [Security model](https://www.assistant-ui.com/generative-frame/docs/security.md): What isolates a widget from the host page, the default Content Security Policy, and what the host still decides.
- [Spec mode](https://www.assistant-ui.com/generative-frame/docs/spec-mode.md): Let the model build UI from your own components with a flat spec streamed as JSONL patches.
- [Theming](https://www.assistant-ui.com/generative-frame/docs/theming.md): How the host page's theme becomes CSS variables in the frame, under canonical and MCP Apps names, and stays in sync.
- [Tools and prompts](https://www.assistant-ui.com/generative-frame/docs/tools.md): The provider-agnostic widget tools, their inputs and results, adapters, and the guidance text the model reads.