Get This Tool
FableCut
Pricing
- Model
- Free
Summary
Most video automation pipelines stall at the same place: the editor hands you a rendered file, but the AI agent has no way to see or touch the timeline itself. FableCut flips that — it exposes the entire timeline as a single JSON document that an agent can read, rewrite, and watch reload in the browser without human intervention.
FableCut is a browser-based, Premiere-style non-linear video editor with zero npm dependencies, designed from the ground up so that AI agents — Claude Code, Claude Desktop, or anything that speaks MCP or REST — can drive the timeline directly. The JSON document *is* the project: agents write to it, the UI reflects the change live. That's the promise. The wall appears when you need effects, color grading, audio mixing, or any of the post-production work that professional editors expect — the docs describe a lean, agent-first tool, not a full-featured studio. Teams that hit that ceiling move to a traditional NLE and use FableCut only for the automated rough-cut stage.
Bottom line: Bet on FableCut when you need an AI agent to assemble a rough cut from a script or structured data and a human to review before export — plan a different tool when that human expects color correction or a plug-in ecosystem on the same timeline.
Community Performance Report Card
No community ratings yet. Be the first to rate this tool!
Community Benchmarks Community
Sign in to submit a benchmarkNo community benchmarks yet. Be the first to share a real-world data point.
Pros
Sign in to edit- The entire timeline is a single JSON document, so an AI agent can read the current cut, modify it, and push changes without screen-scraping or simulated input — which means automation that breaks in every other editor works here by design.
- MCP and REST endpoints are both first-class, so agents built on Claude or any REST-capable framework can control the timeline without writing a custom adapter layer.
- Zero npm dependencies mean deployment is a git clone and a node command — no dependency resolution failures, no supply-chain audit overhead, so the tool runs in environments where package managers are locked down.
- Live-reloading UI reflects agent edits as they happen, so a human reviewer can watch the cut assemble in real time rather than waiting for a render to validate the agent's work.
- MIT license with self-hosted option means the entire system runs on your own infrastructure, so footage and project data never leave your network — which matters the moment a client contract has data residency clauses.
Cons
Sign in to edit- The editor is scoped to cutting: clip assembly, timing, and timeline structure. There is no documented support for color grading, audio mixing, motion graphics, or effects. A team that reaches the rough-cut stage and needs finishing work has to export and re-import into a conventional NLE, which means maintaining two project states and risking sync drift between them.
- The project has ten commits and a small star count at time of curation — community reports and issue history are too thin to confirm stability under sustained agent load or concurrent editing sessions. Teams running unattended overnight agent pipelines have no community precedent to reference, and a production failure has no support channel beyond filing a GitHub issue.
- There is no documented rendering pipeline beyond what the browser's native APIs provide. Teams that need specific codec outputs, broadcast-spec exports, or frame-accurate renders will find the export surface underdocumented — at which point they are evaluating whether FableCut is the editor or just the cut-decision layer in a larger FFmpeg-backed pipeline, which is a different architectural bet than the tool presents.
Community Reviews
Sign in to write a reviewNo reviews yet. Be the first to share your experience.
About
- Platforms
- Browser, Node.js
- API Available
- Yes
- Self-Hosted
- Yes
- Last Updated
- 2026-07-09T19:01:56.831Z
Best For
Who it's for
- AI agent video workflows
- Lightweight browser editing
- JSON-based project control
- Live agent-driven timelines
What it does well
- AI agents autonomously editing video via JSON/MCP/REST
- Browser-based non-linear video editing without installs
- Collaborative human + AI timeline editing
- JSON-driven video project automation
Integrations
Discussion Community
Sign in to commentNo discussion yet. Sign in to start the conversation.
Spotted incorrect or missing data? Join our community of contributors.
Sign Up to ContributeCommunity Notes & Tips Community
Sign in to contributeBe the first to contribute. General notes, observations, gotchas, and tips from people who use this tool day-to-day.
Frequently Asked Questions
- Is FableCut free?
- Yes — FableCut is fully free to use. There is no paid tier.
- Is FableCut open source?
- Yes. FableCut is open source.
- Does FableCut have an API?
- Yes. FableCut exposes a developer API. See the official documentation at https://github.com/ronak-create/fablecut for details.
- Can I self-host FableCut?
- Yes. FableCut supports self-hosting on your own infrastructure.
- What platforms does FableCut support?
- FableCut is available on: Browser, Node.js.
Hours Saved & ROI Stories Community
Sign in to contributeBe the first to contribute. Concrete time/cost savings, with context. e.g. "Cut my code review backlog from 4h to 45m per week."
Curated lists that include this category
Most video editors treat the timeline as a closed box — you get import and export, not direct programmatic access to the cut itself. FableCut opens that box. It is a non-linear video editor that runs entirely in the browser with no install beyond a single Node server, and it exposes every edit decision — clips, cuts, timing — as one JSON document. You edit that document by hand, through the UI, or by pointing an AI agent at the MCP or REST endpoint and letting it rewrite the timeline while you watch the changes appear live.
The differentiating feature is the control surface, not the canvas. Because the entire project state is a JSON document, an AI agent does not simulate editing — it performs it. Claude Code or Claude Desktop can receive a brief, parse it, write clip positions and durations to the document, and the browser reflects the result in real time. The vendor describes this as the whole timeline updating live as the agent works, which means a human reviewer can watch the cut take shape rather than waiting for a rendered export.
FableCut fits tightly into one workflow: automated rough-cut assembly where the logic lives outside the editor. A pipeline that turns a structured script or data feed into a video sequence is well-served here. A team that needs motion graphics, audio sweetening, color grading, or third-party plug-ins is not — the project is MIT-licensed and openly maintained on GitHub, but the docs describe a lean, agent-first tool built for JSON-driven automation, not a replacement for a full production suite. Teams that need those capabilities use FableCut for the agent-driven cut stage and hand off to a conventional NLE for finishing work.
The technical footprint is notable: the vendor states zero npm dependencies and a single Node server entry point, which means deployment is a clone and a node command — no dependency tree to audit, no package manager conflicts. MCP and REST endpoints are both exposed, so agents that speak either protocol can control the timeline without a custom adapter. The project structure on GitHub includes a dedicated MCP server file alongside the main app server, confirming both interfaces are first-class rather than afterthoughts.