Concat Puts Video Editing Under Agent Control

The compelling idea is not another CapCut clone—it is turning a full local video editor into programmable infrastructure.

• –Native Rust and GPU compositing provide a stronger performance foundation than browser-based editors
• –Local Whisper captions, text-to-speech, voice cloning, and background removal keep footage and models on-device
• –JSON-RPC, gRPC, MCP, and CLI access expose the same editing engine to scripts and agents
• –Cross-platform support makes automated editing practical across desktop and mobile workflows
• –It is still beta, so reliability and feature depth have not yet caught up with mature editors