The full upstream README, mirrored here for reference. Install config, tool schemas, adoption signals, and an original overview live on the Promptpilot MCP Server listing page.
Generate images, video, and audio directly in Claude Code, Cursor, Windsurf, or any MCP-compatible AI agent.
20+ models — Flux, GPT-Image-1, Imagen 4, Grok Imagine, Seedance, ElevenLabs TTS, and more. Free models work without an API key. Paid models require a Pollinations key.
With Pollinations API key (for paid models):
Add to ~/.cursor/mcp.json:
Add to ~/.codeium/windsurf/mcp_config.json:
Add to .vscode/mcp.json in your workspace:
| Tool | Description |
|---|---|
list_models | List all image/video/audio models with pricing |
list_styles | List styles, lighting, camera, mood, color presets |
build_prompt | Build optimized prompt from structured inputs |
generate_image | Generate image (returns URL) |
generate_video | Generate video (returns URL) |
generate_audio | Generate speech or music (returns URL) |
check_balance | Check Pollinations API balance |
generate_batch | Generate up to 10 images in one call |
Generate an image:
"Generate a cyberpunk cityscape at night using the flux model"
Batch generation:
"Generate 5 product photos with different backgrounds"
Build + generate:
"Build me a portrait prompt with dramatic lighting and golden hour mood, then generate it"
Check what's available:
"What video models do you have and what are their prices?"
PromptPilot also offers a REST API for programmatic access from any language:
Endpoints:
POST /api/v1/generate — generate image/video/audioGET /api/v1/models — list models with pricesGET /api/v1/balance — check balanceGET /api/v1/schema — OpenAPI schemaGET /api/v1/AGENT_CONTEXT.md — quick reference for agentsMIT