MCP server

Your agent can make the video itself

Solo AI Video ships a Model Context Protocol server. Point Claude Code, Codex or Cursor at it and the agent picks a model, starts the generation, polls it and hands you back the file — without you opening the app.

Three steps

Generate a tokenSettings → API Access → Generate API token. It starts with syn_ and is shown once, so save it as you would any other key.

Point your client at the serverAdd this to your MCP client configuration, with the path to your checkout and the token you just made.

{
  "mcpServers": {
    "solo-ai-video": {
      "command": "node",
      "args": ["/path/to/syntelligence-video/mcp/server.mjs"],
      "env": {
        "SYNTELLIGENCE_API_URL": "https://solo-ai-video.com",
        "SYNTELLIGENCE_TOKEN": "syn_your_token_here"
      }
    }
  }
}

Ask for a videoRestart the client. "Make me an eight-second drone shot over a foggy harbour at dawn" is enough — the agent chooses the model, spends the credits and reports back with the URL.

What the agent gets

list_models: Every model in the lineup, with what it can do and what one generation costs in credits.
generate_video: Start a generation: prompt, model, aspect ratio, resolution, length, sound.
get_generation: Poll a job until it is finished or has failed.
list_projects: Recent projects on the account.
list_templates: The public template library.
get_credits: Credit balance and current plan.

It spends real credits

A generation started by an agent costs exactly what the same generation costs in the app — the price comes from the same place. A failed generation is refunded automatically, and the token can be revoked from Settings at any time.

Get a token and wire it up

Start free