Create and edit videoswith Claude

Model Context Protocol. It's an open standard that lets AI assistants connect to external tools. If your AI supports MCP (Claude, Cursor, Kiro, etc.), it can use Idea to Motion.

Your AI does all the creative work (script, directions, code) using your own LLM tokens. We only spend compute on voice synthesis, validation, and rendering. That's why we can offer unlimited free creation and only charge for final MP4 export.

Any MCP-compatible client: Claude Desktop, Claude Code, Cursor, Kiro, Windsurf, Gemini CLI, VS Code with Copilot, and more. If it can connect to an MCP server, it works.

Your AI calls our MCP tools in sequence: create a job, write a script, generate voiceover, direct scenes, code Remotion components. We handle TTS, validation, assembly, and rendering.

BYOK (Bring Your Own Key): connect your ElevenLabs, Cartesia or Sarvam AI API key and we use it for voice synthesis. 5 curated voices with word-level timing.

Yes. Give your AI a silent screen recording of your product and ask for a product walkthrough. It reads the recording frame by frame on your machine, keeps every step in order, cuts the waiting, and rebuilds each screen as motion graphics with a live cursor and a voiceover. No frame of the original recording appears in the video.

Yes. Tell your AI what to change (e.g. 'make the intro faster'). It edits the scene code directly and the video recomposes in place. You can iterate as many times as you want for free.