The full upstream README, mirrored here for reference. Install config, tool schemas, adoption signals, and an original overview live on the Avots MCP listing page.
Official MCP (Model Context Protocol) server for avots.ai - a multi-provider AI platform.
One connection gives you:
The server lives at https://mcp.avots.ai/ and speaks the MCP 2025-06-18 spec over Streamable HTTP. Tools are billed per call against your avots balance (same balance you'd see on the web app or the Telegram bot).
av_mcp_<48hex>).| Client | Guide | Auth model |
|---|---|---|
| Claude.ai web | docs/claude-web.md | OAuth - paste URL, click Connect, sign in (no token copy-paste) |
| Claude Desktop | docs/claude-desktop.md | Bearer token via mcp-remote |
| Claude Code (CLI) | docs/claude-code.md | Bearer token via claude mcp add |
| Cursor | docs/cursor.md | Bearer token via mcp-remote |
| Cline | docs/cline.md | Bearer token via mcp-remote |
| Any other MCP client | docs/tools.md - endpoint + tool list | Bearer header Authorization: Bearer av_mcp_… |
Ready-to-paste mcp.json snippets live under examples/.
Nineteen tools, all documented in docs/tools.md:
| Tool | Cost | What it does |
|---|---|---|
check_balance | free | Current tokens, subscription tier and the full pricing catalog. |
list_models | free | All active models with per-call cost (filter by chat, image, video, audio). |
list_avatars | free | The user's saved reusable face identities. |
list_trends | free | The avots Studio catalog of viral templates (face-swap videos, ads, animations). |
check_job | free | Poll any async job by job_id. |
chat | ~10-1000 ⚡ | Send a prompt to any chat model. Useful for delegating to GPT, DeepSeek, Sonar, etc. |
generate_image | ~200-500 ⚡ | Sync image gen AND photo editing (image_urls). Recraft can return native vector SVG (format: "svg", 2x). |
generate_video | ~200-5000 ⚡ | Async video gen with two-step confirmation; supports i2v, multi-photo character refs and motion-reference clips. |
face_swap_video | ~500-2000 ⚡ | Swap the face in an existing video with a photo (or saved avatar). Two-step. |
edit_video | ~450 ⚡/sec | Scene edit of an existing clip by text prompt; person, motion and original audio preserved. Two-step. |
lipsync_video | ~300-600 ⚡ | Re-voice an existing talking video: new text (TTS) or a ready audio track. Two-step. |
create_avatar | free / ~200-500 ⚡ | Save a reusable face identity from a photo (free) or generate one from a description. |
generate_talking_avatar | ~600-2500 ⚡ | A portrait speaks your exact text: TTS + lip-sync, quality/fast tiers. Two-step. |
generate_vlog | preview shows price | Vertical AI-influencer clip for Shorts / TikTok: the server writes the line from your topic. Two-step. |
generate_audio | ~50-800 ⚡ | Music (ElevenLabs Music, ACE-Step, Stable Audio) or TTS narration incl. cloned voices. Async. |
create_montage | ~200 ⚡ | Slideshow reel from 4-25 photos: Ken Burns + crossfades + music. |
create_travel_poster | ~200-500 ⚡ | Face photo → vintage travel poster of any country. Synchronous. |
recreate_trend | per trend | Put the user's face into a viral template from list_trends. |
create_calendar_event | ~5 ⚡ | Natural language → event in the linked Apple/Google calendar (or an .ics download). |
About the two-step video flow. Video is the most expensive tool. To avoid surprise spend,
generate_videoreturns a preview card the first time it's called (no submit, no reserve). The client (e.g. Claude) shows the cost + alternative models with prices and asks the user. The user confirms, the client re-calls with the chosenmodel+confirmed: true, and only then does the job get submitted. On submit error the server returns the same alternatives card - it never silently swaps to a pricier model.
A few things this is actually useful for. Each one chains two or more tools through one connection and one balance.
Each of these flows is written out as a runnable script — exact prompt, tool sequence, model picks, cost — in docs/recipes.md.
Cost ranges from ~200⚡ for a single image to ~5000⚡ for a 10-second 1080p Kling Pro clip — run list_models (free) at any time for live per-call prices.
All tool calls bill against your avots balance, just like the web app and Telegram bot. No separate metering. Daily USD cap (set in Settings) and per-key rate limits apply.
See pricing.avots.ai for token packs and subscriptions.
Common cross-client issues (401s, daily caps, two-step video flow, image rendering, npx PATH gotchas) are collected in docs/troubleshooting.md. For client-specific setup, see the per-client guide linked in the table above.
Open an issue here, or write to hello@avots.ai. For platform questions (billing, models, web app) the Telegram bot @AvotsAIbot is the fastest channel.
MIT - feel free to fork the docs, the examples, and anything else here.