ClarifyThe autonomous CRM that helps you sell more
Promoted
OpenChatCut
Open-source AI agent video editor with a real timeline
165 followers
Open-source AI agent video editor with a real timeline
165 followers
OpenChatCut is an open-source ChatCut alternative: local-first AI video editing where Codex, Claude, and MCP agents work on a real multitrack timeline. Free, editable exports, AGPL.
The "real multitrack timeline" part is what makes this stand out to me — most AI video tools spit out a black-box render you can't touch, so having the agent edit an actual editable timeline (and local-first) is a much saner workflow. Love that it's AGPL too.
Quick question: since it works with Claude/Codex/MCP agents — do I bring my own API key/model, or is there a local-model option for people who want it fully offline?
OpenChatCut
@jason_chenyan Thanks—that editable, local-first timeline is exactly the goal. The current setup is BYO key/model: every provider has its own Base URL, API key, and model configuration.
Because OpenAI-compatible Base URLs are configurable, you can point the agent at a local compatible endpoint and use a local model for chat-driven editing. There is not yet a dedicated zero-config local-model/offline switch. Core timeline editing and project data remain local, while optional cloud generation, transcription, stock search, and similar tools still require their respective providers.
Terminal Candy
i cut short product demos for social every week and the timeline is always where ai tools fall apart. they generate clips but you can't nudge anything after. agents on a real multitrack timeline over mcp is the missing piece. can claude code drive it today or do i need your chat ui?
OpenChatCut
@terminal_candy Yes—Claude Code can drive OpenChatCut through MCP today; the built-in chat UI is not required. Once the target project is open in OpenChatCut, MCP exposes the live timeline and editing tools to the external agent.
The current caveat is that external MCP edits apply directly to the working timeline, so they rely on the normal undo/redo and named-version workflow. The propose/review/apply layer currently belongs to the built-in chat agent.
Terminal Candy
@0xsline perfect, that answers it. edits landing on the real timeline with normal undo is exactly what i was missing. going to wire claude code into it this week and run it through my weekly cuts. thanks for the detailed reply.
The real-timeline part is what sells me — chat-only editors feel like a black box, but coming from a DAW world, being able to drop into a multitrack timeline and hand-nudge things after the agent's pass is exactly the workflow I trust. Quick one: how deep does the audio side go — multitrack audio with volume keyframes/ducking, or mostly video-first for now? Local-first + AGPL is a great call btw.
OpenChatCut
@lennoxbeflying The audio side is already more than a placeholder: OpenChatCut supports multiple audio tracks, per-clip volume and fades, track mute/lock controls, voice recording, loudness normalization, voice isolation, and role-based automatic ducking with adjustable duck depth.
What it does not have yet is a DAW-style volume automation/keyframe lane. Audio volume is currently static per clip, plus fades and automatic ducking. So it is deeper than a video-first editor, but not yet a full DAW.
回复 Simon
Congrats on the launch. The real timeline angle is the part that makes this feel useful, especially compared with one-shot AI video tools. When an agent proposes edits across clips, captions, and transitions, is there a clear diff/review step before applying changes? I would love to know how easy it is to accept one cut, reject another, and still keep manual timeline control.
OpenChatCut
@simon_dao Yes. In the built-in agent flow, edits are first executed against a scratch project document rather than the committed timeline. The proposal card lists each operation with its action, target, and impact, and lets you preview the result, deselect individual operations, reject everything, or apply only the selected changes.
The selected edits are then applied as one undoable change. If the timeline changed after the proposal was created, OpenChatCut marks it as stale and asks you to re-propose or explicitly force-apply it. Manual timeline control remains available throughout.
One current distinction: edits coming directly from an external MCP agent bypass this proposal layer and modify the working timeline directly.
finally something that doesnt lock your timeline behind a subscription, ran a quick cut with claude and it actually respected my existing tracks instead of nuking them
OpenChatCut
@canertaanux3r That’s exactly the experience we’re aiming for — your timeline should stay yours, and AI should work with your existing tracks instead of flattening or rewriting them. Glad Claude handled the quick cut cleanly. Thanks for giving it a try!
Love the local-first approach and the multitrack timeline is exactly what I was missing. One thing that would really help: add a visual diff or changelog view when the AI agents make edits, so I can quickly review what Codex or Claude changed before approving. Right now trusting invisible agents on a timeline feels a bit opaque.
OpenChatCut
@lknurhallulrfs Thanks—this is exactly the kind of transparency we care about. The built-in agent is not entirely a black box today: edits are first executed against a scratch project document, and the proposal card lists each operation’s action, target, and impact. Users can preview the result, deselect individual operations, reject everything, or apply only the selected changes. A proposal is also marked as stale if the timeline has changed.
That said, a full visual before/after diff and persistent changelog are not implemented yet. Edits made by external Codex or Claude agents through MCP also modify the working timeline directly. This is a real transparency gap in the current version.
Love the local-first approach and editable exports being free. One thing that would really help adoption is a built-in shareable project link feature where collaborators can open the same timeline in a read-only browser view without installing anything, perfect for getting quick feedback from clients or teammates before they commit to the setup.
OpenChatCut
@bnyaminzopwhd4 Thanks! Projects, media, and chat history are stored locally, and a project can be exported as a portable .ccproj.json file and imported into another OpenChatCut instance.
There is not yet a built-in public sharing link or a zero-install, read-only browser view. Local project URLs only work within the same local service and storage environment, so someone receiving a project file still needs to run OpenChatCut. The read-only review mode you described would be a very useful collaboration layer while preserving the local-first model.