generate_image runs Flux2 Klein, Nano Banana 2, Nano Banana Pro or GPT Image II. generate_video runs Google Veo 3.1 and LIVS 2.0. restyle_video runs the LIVS PRO EDIT pipeline over your own footage. job_status and jobs_wait follow renders to completion, show_generations reads your history, list_models and balance keep the agent inside your plan, and upload_media stores reusable references.
Model Context Protocol
Connect Livs to Claude, Codex and any MCP client
Livs ships a remote MCP server at api-gen.livs.es/mcp. Point Claude Code, Claude Desktop, Codex or Cursor at it with a Livs API key and your agent can generate images and videos, track render jobs and read your gallery, billed against the same Livs credits you use in the web app.
Start with free credits
- Remote MCP server over Streamable HTTP
- Claude Code, Claude Desktop, Codex and Cursor
- Image and video generation tools
- Bearer API keys with scopes
- Shared credits and usage dashboard
The server speaks MCP Streamable HTTP with bearer authentication, so Claude Code connects with a single claude mcp add command and Claude Desktop with a small mcpServers block. Codex and Cursor connect through the @livs/mcp npm bridge. Keys are created and revoked from your Livs workspace and every tool call is logged against your account.
Agent generations cost exactly what they cost in the app: 6 credits per image and 35 per video. Video generation requires a paid plan. Credits are charged when a job is created and refunded when a render fails for technical reasons.
It connects a premium interface, AI media generation, billing, user accounts and downloadable galleries so creators can return every week and keep producing.
https://api-gen.livs.es/mcp. Authenticate with an Authorization: Bearer header carrying a Livs API key.
Run claude mcp add --transport http livs https://api-gen.livs.es/mcp --header "Authorization: Bearer YOUR_KEY" after creating a key in your Livs workspace.
Yes. Add the @livs/mcp bridge to ~/.codex/config.toml with your LIVS_API_KEY and Codex gets the same tools over stdio.
The same as in the app: 6 credits per image and 35 credits per video. Video generation needs a paid plan.