Arcframe MCP Server

Generate videos, images, audio, and 3D models from any MCP-compatible AI agent — Claude, Cursor, ChatGPT, and more.

Hosted endpoint: `https://mcp.arcframe.ai/mcp`

Auth: OAuth 2.1 (one-click connect — no API key to copy)

Sign up: arcframe.ai

Connect in 60 seconds

Claude Desktop

Add to your `claude_desktop_config.json`:

{
  "mcpServers": {
    "arcframe": {
      "type": "http",
      "url": "https://mcp.arcframe.ai/mcp"
    }
  }
}

Restart Claude Desktop → it will prompt you to sign in to Arcframe via OAuth.

Cursor

Settings → MCP → Add new server → paste `https://mcp.arcframe.ai/mcp`.

Other MCP clients

Any client that supports Streamable HTTP transport works. Point it at `https://mcp.arcframe.ai/mcp`.

What you can do

Generation is asynchronous: `generate_*` returns a `task_id` immediately. Poll with `get_job(task_id)`, then call `get_job_display(task_id)` to render the result inline.

Models

Video — Veo 3.1, Seedance 2.0, Kling 2.5, Wan 2.5, Sora 2

Image — Flux Pro Ultra, Flux Dev, Imagen 4, Ideogram v3, GPT Image 1.5

Audio — ElevenLabs v3, MiniMax Speech HD, Gemini Flash TTS, Stable Audio 2.5, Mirelo SFX

3D — Trellis 2, Tripo3D v2.5, Rodin v2, Meshy v6, Hunyuan 3D

Call `list_models` from your agent to see live credit costs per model.

Credits

1 credit ≈ $0.05. Free plan includes image generation (Flux Schnell). Video, audio, and 3D require Starter plan or above.

See arcframe.ai/pricing for plan details.

Example prompts

Generate a 5-second cinematic video of a sunset over the ocean using Veo 3.1

Create an image of a futuristic city at night, Flux Pro Ultra

Generate a Hindi voiceover for: "Welcome to Arcframe, your AI creative studio"

Build a 3D model of a wooden chair from this image [attach photo]

Support

• Docs & pricing: arcframe.ai
• Email: support@arcframe.ai