OpenVideoKit

A scene-based, AI-first video editor. A built-in copilot proposes edits you accept or reject, and any MCP agent can drive it.

Try it in 30 seconds

Bundled SPA, 25 templates, edge-tts voiceover, MP4 export, AI copilot, MCP endpoint.

bash

python3 -m venv .venv
source .venv/bin/activate
pip install openvideokit
ovk serve                       # → http://localhost:8000

A walkthrough of the editor, the template catalog, and an end-to-end render.

The editor showing the AI prompt that generated the demo video above.

🇺🇸 Ava Multilingual → 🇫🇷 Vivienne → 🇩🇪 Florian → 🇯🇵 Nanami → 🇻🇳 Hoài My → 🇨🇳 Xiaoxiao → 🇧🇷 Thalita → 🇺🇸 Ava Multilingual.

Live video previews, field defaults, and HTML source for every template — searchable and filterable.

AI Copilot
A LangGraph agent proposes slide edits. Accept or reject each one — ⌘Z works on AI edits like your own.
MCP for any agent
Claude Code, Cursor, Gemini CLI, Goose connect to the same 28-tool API the copilot uses.
322 neural voices
edge-tts across 142 locales. Word-level karaoke captions baked from real TTS timings.
Bring your own key
OpenAI, OpenRouter, Ollama, vLLM, LM Studio. Inference runs locally; the cloud never charges.
25 curated templates →
Titles, charts, kinetic typography, SVG diagrams. Both humans and AI apply them in one call.
WYSIWYG by construction
Preview and MP4 export share one HyperFrames engine. Export surprises are structurally impossible.

Titles, charts, kinetic typography, SVG diagrams — 25 curated layouts with live video previews, field defaults, and copy-ready HTML source.