MCP Install Guide
Playcut ships an MCP (Model Context Protocol) server so you can generate images and videos directly from AI assistants.
Prerequisites
Section titled “Prerequisites”- Node.js 18+
- A Playcut API key (get one here)
Installation
Section titled “Installation”Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):
{ "mcpServers": { "playcut": { "command": "npx", "args": ["-y", "@playcut/mcp-server"], "env": { "PLAYCUT_API_KEY": "pk_live_your_key_here" } } }}Restart Claude Desktop. Playcut tools appear automatically in new conversations.
claude mcp add playcut \ --command "npx -y @playcut/mcp-server" \ --env PLAYCUT_API_KEY=pk_live_your_key_hereOr add to your project’s .mcp.json:
{ "mcpServers": { "playcut": { "command": "npx", "args": ["-y", "@playcut/mcp-server"], "env": { "PLAYCUT_API_KEY": "pk_live_your_key_here" } } }}In Cursor settings → MCP, add a new server:
- Name:
playcut - Command:
npx -y @playcut/mcp-server - Env:
PLAYCUT_API_KEY=pk_live_your_key_here
Available tools
Section titled “Available tools”| Tool | Description |
|---|---|
generate-image-from-text | Text → image |
generate-image-from-reference | Reference images → new image |
generate-video-from-text | Text → video |
generate-video-from-image | Animate a still image |
generate-video-interpolation | Interpolate between two frames |
generate-video-extension | Extend a video clip |
generate-tts | Text-to-speech |
design-voice | Create a custom voice from description |
clone-voice-from-audio | Clone a voice from an audio sample |
generate-actor | Generate an AI actor appearance |
actor-shoot | Photo shoot for an actor |
get-task-status | Check task completion |
wait-for-task | Block until task completes |
list-assets | Browse your media library |
download-asset | Get a signed download URL |
Example
Section titled “Example”After connecting, ask Claude:
“Generate a 16:9 image of a futuristic city at night, photorealistic”
Claude calls generate-image-from-text, waits for completion, and shows you the result.