realtimeavatar.ai is built with React, hosted on Cloudflare. It measures traffic with Cloudflare Web Analytics.
Built with React and TanStack Start, hosted on Cloudflare. Domain registered in 2026.
| Technology | Version | First seen |
|---|---|---|
| JavaScript Frameworks & Libraries2 | ||
| – | 9 Nov 2026 | |
| – | 9 Nov 2026 | |
| Frontend Build Tools & CSS3 | ||
| – | 9 Nov 2026 | |
| CDN & Hosting1 | ||
| – | 9 Nov 2026 | |
| Analytics & Tag Management1 | ||
| – | 9 Nov 2026 | |
| Video, Audio & Media Delivery1 | ||
| – | 9 Nov 2026 | |
Registration from the registry's RDAP record, the certificate the site serves, and its rank in public traffic lists.
The crew hasn’t read this site’s colors yet. We read a palette when we capture a site’s screenshot, busiest sites first. Meanwhile, browse palettes from other sites.
route2.mx.cloudflare.netroute3.mx.cloudflare.netroute1.mx.cloudflare.netishaan.ns.cloudflare.comperla.ns.cloudflare.comgoogle-site-verification=0vETD5qVtdAmLkgOIilTc9G-NgKy-CRoz1_0lO8zWcYv=spf1 include:_spf.mx.cloudflare.net ~allv=DMARC1; p=reject;# Realtime Avatar — crawlers and AI agents welcome. # Machine-readable surface: # /llms.txt — agent guide to the platform and API # /llms-full.txt — agent guide plus the full English text of every doc page and post # (translations are linked from /llms.txt, not inlined here) # /openapi.json — OpenAPI 3.1 spec for /api/v1 # MCP server — `npx -y realtime-avatar-mcp` (npm, stdio). Reads the account so an # agent does not have to guess avatar ids. https://realtimeavatar.ai/docs/mcp # /api/realtime-avatar — same-origin LiveKit session grant proxy # /feed.xml — RSS feed of the blog # /blog/<slug>.md — every blog post as raw markdown (append .md) # /docs — developer docs; every page also at <page>.md # /.well-known/oauth-protected-resource/api # — RFC 9728: the bearer scheme and every scope a key can carry # # Any page also answers `Accept: text/markdown` with markdown at its own URL. User-agent: * Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ # ...except the public asset tier, which is where every social card lives. # # Each blog post's og:image is served from /api/assets/public/characters/<slug>/portrait.png or # /api/assets/public/blog/<slug>/hero.webp, so `Disallow: /api/` was telling Facebook, LinkedIn and # X not to fetch the one image the card is made of. Those crawlers honour robots.txt, and they fall # under `*` — the AI-agent groups below never had the Disallow, so this was invisible to any check # that only asked whether the image URL resolves. It does resolve; the crawler was told not to ask. # # A longer path wins over a shorter one, so this reopens exactly the asset tier and nothing else: # /api/v1, the proxy, and every other endpoint stay disallowed. The tier is unauthenticated by # design — it is what the site itself serves portraits from. Allow: /api/assets/public/ # AI/LLM crawlers explicitly allowed on public content. # # Each group below repeats the `*` rules in full, and that repetition is required rather than # untidy: a crawler that matches a named group uses ONLY that group and never reads `*`. These # groups used to carry `Disallow: /platform/dashboard` alone, which silently granted every named # agent the three areas `*` protects — while the header called them public content. The same # divergence is what made the social-card bug above invisible. # # `robots-groups-agree.test.ts` holds them equal; drift fails there rather than here. User-agent: GPTBot Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ User-agent: OAI-SearchBot Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ User-agent: ChatGPT-User Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ User-agent: ClaudeBot Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ User-agent: Claude-User Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ User-agent: Claude-SearchBot Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ User-agent: PerplexityBot Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ User-agent: Perplexity-User Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ User-agent: Google-Extended Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ User-agent: Applebot-Extended Allow: / Disallow: /platform/dashboard Disallow: /platform/billing Disallow: /platform/settings Disallow: /api/ Allow: /api/assets/public/ Sitemap: https://realtimeavatar.ai/sitemap.xml
# Realtime Avatar
> A realtime AI avatar API and TypeScript SDK. Give a character a face and a voice, put them
> on a live call in a browser or a native app, and let them use your tools. One mint endpoint
> on your server decides who may call and what the character knows; one component renders it.
The sections between generated markers below are rebuilt by
`bun scripts/content/sync-content.ts` — do not hand-edit them.
## Start here
- [Realtime Avatar](https://realtimeavatar.ai/): the product itself — what it is, what it replaces, and a prompt you can hand to a coding agent. Also as markdown at https://realtimeavatar.ai/index.md, which inlines the whole FAQ.
- [Pricing](https://realtimeavatar.ai/pricing): every plan and what a live minute costs. Also as markdown at https://realtimeavatar.ai/pricing.md
- [Overview](https://realtimeavatar.ai/docs): what the product is and what a call costs.
- [Quickstart](https://realtimeavatar.ai/docs/quickstart): key → app on an example avatar → your own character.
- [Authentication](https://realtimeavatar.ai/docs/authentication): what belongs on your server and why.
## Agent tools (MCP)
If you are a coding agent, you do not have to guess at this account. There is an MCP server —
`realtime-avatar-mcp` on npm, stdio, no build step — that tells you which avatars actually
exist, what a call costs, and what the balance is, before you write any code. Avatar ids
cannot be guessed; call `list_avatars` first.
{
"mcpServers": {
"realtime-avatar": {
"command": "npx",
"args": ["-y", "realtime-avatar-mcp"],
"env": { "REALTIME_AVATAR_API_KEY": "YOUR_API_KEY" }
}
}
}
Read tools (`list_avatars`, `get_avatar`, `credit_balance`, `list_sessions`, `list_clips`) are
on by default. Set `REALTIME_AVATAR_ALLOW_WRITES=1` to also get `create_avatar_from_image`,
`create_avatar_from_video`, `upload_asset`, `create_remote_asset`, `start_call`, `set_loop`,
`set_clip_library` and `sync_clips`. Full page: https://realtimeavatar.ai/docs/mcp.md
The server is a BUILD-TIME tool. Applications should still mint sessions from their own
server route — that endpoint is where the caller is authorised and the character briefed.
## Machine-readable surfaces
- MCP server for agents: `npx -y realtime-avatar-mcp` — see above, and https://realtimeavatar.ai/docs/mcp.md
- Full docs as one file: https://realtimeavatar.ai/llms-full.txt
- Every docs page as markdown: append `.md` to its URL.
- Or ask any page for markdown at its own URL: send `Accept: text/markdown`. Pages with a
published mirror answer `200 text/markdown` and name it in `Content-Location`; pages without
one answer HTML if you will take it, and `406` if you said you would not.
- The landing page as markdown, FAQ inlined: https://realtimeavatar.ai/index.md
- Pricing as markdown: https://realtimeavatar.ai/pricing.md
- OpenAPI 3.1 for the whole API: https://realtimeavatar.ai/openapi.json, 30 operations,
request and response schemas, and the scope each one needs, machine-readably in
`x-required-scopes` on every operation and `x-scopes` on the security scheme. Generate a
client from it rather than reading the endpoint table.
- API authentication, as RFC 9728 protected-resource metadata:
https://realtimeavatar.ai/.well-known/oauth-protected-resource/api — the bearer scheme and
every scope a key can carry. There is no OAuth authorization server: keys are minted in the
dashboard or with `POST /api/v1/api-keys`, and a 401 or 403 says so in `WWW-Authenticate`.
- A wrong path answers a real `404` everywhere, with the site map in the body — as markdown or
JSON if you asked for either, and always as JSON under `/api`.
<!-- generated:pricing:start -->
## Pricing
One meter: seconds on air. Speech recognition, the LLM, the voice and rendering are included. Published self-serve overage is $0.08–$0.095 per minute ($4.80–$5.70 per hour). Your own hosting and external tools remain separate.
| Plan | Monthly | Included realtime | Overage | Concurrent streams |
| --- | --- | --- | --- | --- |
| Sandbox | Free | 8 min | — | 2 |
| Starter | $9/mo | 120 min | $5.70/hour | 3 |
| Developer | $24/mo | 600 min | $5.10/hour | 10 |
| Studio | $119/mo | 3,000 min | $4.80/hour | 20 |
| Scale | Quoted | Your volume | Quoted | Custom |
Creating a character is billed separately from talking to it: $3 per avatar beyond the number your plan includes, its generated resting loop included. Credits are one pot: the same balance pays for generating a character and for the minutes it then talks.
Standalone speech, a workspace voice turning text into an audio file outside a call, costs 1 credit per 90 bytes of text, rounded up, with a minimum of 1 credit per request. Plain English is one byte per character; accented letters and other scripts use 2 to 4. It is open to workspaces on a paid plan or that have bought credits.
A pre-rendered video, an avatar speaking a script to an MP4 with no live call, costs 1 credit per second of rendered video, the same rate as a live call, up to 180 seconds per video. A render that fails is not charged.
Scale is priced by conversation rather than published, so no figure for it appears on this site.
Full table: https://realtimeavatar.ai/pricing · as markdown: https://realtimeavatar.ai/pricing.md
Cost calculator: https://realtimeavatar.ai/tools/avatar-cost-calculator · worked scenarios: https://realtimeavatar.ai/blog/realtime-ai-avatar-api-pricing
<!-- generated:pricing:end -->
<!-- generated:docs:start -->
## Documentation (raw markdown)
Every docs page is mirrored as plain markdown — append `.md` to any docs URL.
- [Realtime Avatar API](https://realtimeavatar.ai/docs.md): Put a live character in your product, on a voice or video call, with your own tools wired into the conversation.
- [Quickstart](https://realtimeavatar.ai/docs/quickstart.md): From an API key to a live avatar call in three steps: get a key, build the app on an example avatar, then swap in a character of your own.
- [Authentication](https://realtimeavatar.ai/docs/authentication.md): Get a key, keep it on your server, and decide who may start a call.
- [Build with Lovable, Bolt, v0 and Replit](https://realtimeavatar.ai/docs/ai-app-builders.md): Add a live AI avatar call to an app built in Lovable, Bolt.new, v0 or Replit: a copy-paste prompt per builder, where the API key goes, and a starter repo that opens in one click.
- [Build a realtime AI avatar in Next.js](https://realtimeavatar.ai/docs/nextjs.md): Build a working Next.js avatar app with a server-side session proxy, React video UI, a complete local starter, and troubleshooting.
- [TanStack Start](https://realtimeavatar.ai/docs/tanstack-start.md): Mount the realtime avatar proxy in a TanStack Start app: a splat server route on the server, one component in the browser.
- [Express](https://realtimeavatar.ai/docs/express.md): Mount the realtime avatar proxy as Express middleware — Express 4 or 5 — with the same authorize and session policy every adapter takes.
- [Hono, Workers, Bun and Deno](https://realtimeavatar.ai/docs/hono.md): Mount the realtime avatar proxy on any Fetch-handler runtime — Hono, Cloudflare Workers, Bun or Deno — including the apiKey factory Workers require.
- [Realtime AI avatars with React and React Native](https://realtimeavatar.ai/docs/react.md): Embed live AI avatar calls in React or React Native with AvatarCall, useAvatarCall, connection states, and a server-side session proxy.
- [Build avatar apps with the Realtime Avatar MCP server](https://realtimeavatar.ai/docs/mcp.md): Connect a coding agent to Realtime Avatar: configure the MCP server, inspect available avatars, and build a working app with the TypeScript SDK.
- [Creating an avatar](https://realtimeavatar.ai/docs/video.md): One image in, a moving character out: generated motion, idle variations and requestable actions, full-library declarations, loop editing, and generation prices.
- [Calls](https://realtimeavatar.ai/docs/sessions.md): What your server decides, what the client reports, and how to end a call without cutting the character off mid-sentence.
- [Editing a character](https://realtimeavatar.ai/docs/editing.md): Change an avatar's clips and resting loop after creation: how to declare them, how each change settles, and the refusals worth telling apart.
- [Tool calling](https://realtimeavatar.ai/docs/tool-calling.md): The platform never executes your tools. Here is the boundary, the client tool plane, and the server loop that turns a call into a working agent with a face.
- [API reference](https://realtimeavatar.ai/docs/api-reference.md): Every public endpoint on https://realtimeavatar.ai/api/v1 — what it takes, what it returns, which scope it needs, and how it fails.
- [Experimental features](https://realtimeavatar.ai/docs/experimental.md): What is experimental right now, what that word promises, and how to turn each one on. Included with Studio and above.
<!-- generated:docs:end -->
<!-- generated:journal:start -->
## Blog (raw markdown)
Published posts as markdown; append `.md` to any post URL.
- [Engineering Live Presence: Building Realtime AI Avatars with Code](https://realtimeavatar.ai/blog/engineering-live-presence-realtime-ai-avatars.md): Explore how realtime AI avatars combine visual presence with intelligent conversation, from a single portrait to interactive voice and tool\-calling\.
- [Architecting AI Roleplay Apps with Persistent Avatar Memory](https://realtimeavatar.ai/blog/architecting-ai-roleplay-apps-with-persistent-avatar-memory.md): Unlock deeper immersion in AI roleplay\. Learn to build applications that give avatars robust, persistent memory across sessions using the Realtime Avatar SDK\.
- [Embodied Narratives: Realtime AI Avatars for Interactive Storytelling Beyond Static Video](https://realtimeavatar.ai/blog/embodied-narratives-realtime-ai-avatars-for-interactive-storytelling.md): Explore how Realtime Avatar empowers developers to build interactive stories with live video avatars, offering a dynamic alternative to static AI video\.
- [Building Interactive Narratives: Realtime AI Avatars for Dynamic Story Generation](https://realtimeavatar.ai/blog/realtime-ai-avatars-for-interactive-story-generation.md): Explore how realtime AI avatars move beyond static video generation, enabling dynamic, interactive stories driven by live conversation and user input\.
- [Build an Avatar Assistant People Can Use Without Speaking or Hearing](https://realtimeavatar.ai/blog/accessible-avatar-assistant-without-voice.md): Build a text\-first appointment assistant with optional avatar calls, verified confirmations, keyboard controls, and clear timeout recovery\.
- [Build an AI Avatar Fantasy Adventure with Choices Beyond a Menu](https://realtimeavatar.ai/blog/ai-avatar-fantasy-adventure-story-game.md): Build a fantasy adventure where players invent solutions, cross a floating market with an avatar guide, and reach an ending shaped by their ideas\.
- [Build an AI Avatar Horror Hotline Players Can Hang Up On](https://realtimeavatar.ai/blog/ai-avatar-horror-hotline-game.md): Build a playable AI avatar horror hotline with a stranded caller, a fictional island map, three consequential choices, and endings players can replay\.
- [Build an AI Avatar Interrogation Mystery Players Can Solve](https://realtimeavatar.ai/blog/ai-avatar-interrogation-mystery-game.md): Build a playable AI avatar mystery with three suspects, cross\-checkable clues, a fixed solution, and an accusation that rewards evidence over endless chat\.
- [Build an AI Avatar Party Game with an Improv Host](https://realtimeavatar.ai/blog/ai-avatar-party-game-improv-host.md): Build a three\-round AI avatar party game where friends pitch absurd inventions, survive a twist, and vote for a winner, with a host who keeps it moving\.
- [The $1M ARR AI Companion Playbook for 2027](https://realtimeavatar.ai/blog/ai-companion-million-arr-playbook-2027.md): Build toward a $1M ARR AI companion business: choose a niche, rent realtime infrastructure, cap call costs, and test retention and distribution\.
- [Build an AI Customer Support Avatar That Knows When to Hand Off](https://realtimeavatar.ai/blog/ai-customer-support-avatar-human-handoff.md): Build a support avatar with authenticated order lookup, consent\-based summaries, durable tickets, and honest human\-handoff states using Realtime Avatar\.
- [Build an AI Email Assistant That Shows the Evidence Behind Its Briefing](https://realtimeavatar.ai/blog/ai-email-assistant-source-linked-briefings.md): Build a read\-only AI email briefing with source\-linked exceptions, visible sync limits, thread comparisons, and optional voice discussion before a trip\.
- [Build an AI Game NPC That Respects Quest State and Spoilers](https://realtimeavatar.ai/blog/ai-game-npc-dialogue-quest-state.md): Build an AI game NPC with game\-owned quest state, spoiler\-safe context, retry\-safe rewards, save recovery, and optional browser avatar conversations\.
- [Build an AI Museum Avatar Guide That Can Show Its Sources](https://realtimeavatar.ai/blog/ai-museum-avatar-guide-collection-sources.md): Build an AI museum avatar guide with official collection records, visible sources, careful attribution, object\-level image rights, and honest unknowns\.
- [Build AI Roleplay Chat with Edits, Rewinds and Separate Story Branches](https://realtimeavatar.ai/blog/ai-roleplay-chat-edit-rewind-branches.md): Build AI roleplay chat with distinct edit, regenerate, rewind and branch actions, consistent checkpoints, and guards against canceled or stale replies\.
- [Build an AI Shopping Avatar That Compares Products Before Checkout](https://realtimeavatar.ai/blog/ai-shopping-avatar-product-comparison.md): Build a read\-only shopping avatar with exact Shopify variants, source\-backed comparisons, honest stock labels, and price rechecks before checkout\.
- [Build an AI Story Generator That Can Stop, Save and Continue Cleanly](https://realtimeavatar.ai/blog/ai-story-generator-streaming-save-cancel.md): Build a streaming AI story generator with separate drafts, explicit acceptance, reliable stop controls, revision checks and honest recovery after disconnects\.
- [AI Story Generator or Character Roleplay? Choose How You Want to Play](https://realtimeavatar.ai/blog/ai-story-generator-vs-character-roleplay.md): Choose an AI story generator, character roleplay or branching fiction\. Try three cozy prompts, fix unwanted narration and keep your story going\.
- [How to Build an AI Bible Study Chatbot with Voice and an Avatar](https://realtimeavatar.ai/blog/build-ai-bible-study-avatar.md): Build a voice\-and\-avatar Bible study guide with source\-linked passages, clear AI disclosure, translation labels, and application\-owned conversation context\.
- [Build AI Character Video Keepsakes: Turn One Great Scene Into a Takeaway](https://realtimeavatar.ai/blog/build-ai-character-video-keepsakes.md): Build AI character video keepsakes from player\-approved moments, with editable prompts, clear costs, private previews and a small paid\-demand test\.
- [Build an AI Companion App: Make the First Conversation Worth Returning To](https://realtimeavatar.ai/blog/build-ai-companion-app-first-session.md): Build an AI companion app's first session with character choice, a useful opening, optional calls, explicit guest saves, and honest credit exits\.
- [Build an AI Roleplay Comic Generator: Give Players a Story Worth Sharing](https://realtimeavatar.ai/blog/build-ai-roleplay-comic-generator.md): Turn an original roleplay scene into a four\-panel comic with editable storyboards, consistent characters, panel retries, and a paid\-issue experiment\.
- [How to Build an AI Tarot Chatbot with Voice and a Live Avatar](https://realtimeavatar.ai/blog/build-ai-tarot-chatbot-avatar.md): Build an AI tarot chatbot with stable card draws, tool\-backed context, voice and avatar sessions, clear pricing, and a separate astrology calculation layer\.
- [Character\.AI Alternatives for Builders: Consumer App, Agent Framework or Avatar API?](https://realtimeavatar.ai/blog/character-ai-alternative-api-build-vs-buy.md): Choose a Character\.AI alternative for your own app by separating character authoring, agent infrastructure and managed avatar calls, with a concrete test\.
- [Free AI Character Chat: Check the Limits Before Starting a Story](https://realtimeavatar.ai/blog/free-ai-character-chat-voice-video-limits.md): Find free AI character chat that fits your routine\. Check text limits, premium models, ads, voice minutes, video credits, and what happens after a trial\.
- [Build an Interactive Romance Short Drama Players Can Step Into](https://realtimeavatar.ai/blog/interactive-romance-avatar-short-drama.md): Build a replayable romance episode with an AI avatar: a station reunion, meaningful choices, satisfying endings, and a practical live\-call implementation\.
- [MCP, A2A, PAP and Agent Payments: What Each Layer Actually Does](https://realtimeavatar.ai/blog/mcp-a2a-pap-agent-payment-protocols.md): Compare MCP, A2A, PAP, ACP and AP2 by responsibility and maturity, with a shopping\-agent example separating access, approval, payment and verified results\.
- [Keep Personal Agent Tasks Running After the Call Ends](https://realtimeavatar.ai/blog/personal-agent-background-tasks-after-call.md): Keep personal agent tasks running after a voice call with app\-owned jobs, honest status, safe retries, and recoverable context in Realtime Avatar\.
- [Designing a Personal Agent Users Can Understand, Correct and Approve](https://realtimeavatar.ai/blog/personal-agent-trust-approval-interface.md): Design a personal agent approval and correction interface that separates proposals, permission and verified results, with practical Realtime Avatar boundaries\.
- [Build a Personal AI Assistant for Daily Planning with Voice and an Avatar](https://realtimeavatar.ai/blog/personal-ai-assistant-daily-planning-avatar.md): Build a daily\-planning personal AI assistant with calendar context, voice and an avatar\. Keep schedules, memory and confirmed actions in your application\.
- [Personal AI Assistant Memory: Corrections, Provenance and Migration](https://realtimeavatar.ai/blog/personal-ai-assistant-memory-correction-migration.md): Build personal AI assistant memory that preserves corrections, tracks provenance, imports only supported context, and stays accurate across new calls\.
- [Build a Personal Assistant That Continues from Phone to Desktop](https://realtimeavatar.ai/blog/personal-assistant-phone-to-desktop-handoff.md): Design a phone\-to\-desktop assistant handoff with confirmed trip context, authenticated resume, fresh calls, and safeguards against stale device commands\.
- [Build a Talking Virtual Pet Players Want to Visit, Not Worry About](https://realtimeavatar.ai/blog/talking-virtual-pet-avatar-game.md): Build a talking virtual pet around tiny expeditions, meaningful choices and a personal scrapbook, with satisfying visits and no missed\-day guilt\.
- [Building Interactive Avatar Training Courses with Realtime AI](https://realtimeavatar.ai/blog/building-interactive-avatar-training-courses-with-realtime-ai.md): Explore the engineering of dynamic, real\-time AI avatar training courses, moving beyond static video to interactive, adaptive learning experiences\.
- [Your personal agent deserves a face](https://realtimeavatar.ai/blog/give-your-personal-agent-a-face.md): Instinct, Muse, and Tab are making personal agents part of everyday life\. Realtime Avatar helps developers give their own agent a face people can talk to\.
- [From Portrait to Presence: Realtime AI Avatars for Dynamic Training](https://realtimeavatar.ai/blog/from-portrait-to-presence-ai-avatars-for-realtime-training.md): Explore how to build interactive AI training experiences using realtime avatars, created from a single image and deployed for live, dynamic learning\.
- [Field Notes: Realtime AI Avatars for Engaging and Scala| – |
| 9 Nov 2026 |
| – | 9 Nov 2026 |