Most "AI video editors" share a dirty secret: they generate a finished clip and then strand you. Want to nudge a single cut, swap a caption, or fix the pacing? You can't — you're stuck with whatever the model spat out. ChatCut, which launched its latest version on Product Hunt in July 2026, is built specifically to break that pattern. It pairs an autonomous AI editing agent with a real, multi-track timeline you can touch at any point.
Founded by documentary and commercial makers Kaiwen Li and Alima Strickland, ChatCut treats AI as an assistant editor rather than a vending machine. You describe the intent, the agent does the rough cut, and you keep control of the final story. The launch landed at #5 on Product Hunt's daily ranking — a real signal in a category this crowded.
How the editing actually works
You upload footage and talk to it in plain English: cut the silences, build a 30-second highlight, add captions, drop in B-roll. The agent runs the edit while you watch, then hands back a fully editable timeline instead of a flattened export.
For talking-head and interview footage, ChatCut leans on transcription, organizing sound bites into a coherent structure you can then direct. For visually driven material, it analyzes the footage and assembles sequences you can rearrange. Either way, every action is undoable step by step, and you can remap keyboard shortcuts to match Premiere, DaVinci Resolve, or Final Cut.
The design bet, in the founders' own words: AI that gets you 80% of the way while keeping you in control beats AI that promises 90% as an uneditable black box.
It's also a genuine all-in-one. Built-in generation covers video and images — via models like Seedance 2.0, Kling 3.0, GPT Image 2, and Nano Banana — plus ElevenLabs voices, music, transitions, and motion graphics. Crucially, the motion-graphics elements stay editable rather than baked in, and when you're done, XML export hands the project off to a professional editor like Premiere or DaVinci.
The reusable-skills trick
One feature deserves a callout: you can save an editing workflow or a design style as a reusable "skill." Do the same kind of edit repeatedly — say, turning raw interviews into vertical social reels — and you can replay that recipe on new footage instead of starting from scratch each time. It's a smart answer to the biggest weakness of prompt-driven tools, which is staying consistent across projects.
How it stacks up
The AI-editing field is busy, and ChatCut's differentiation is easiest to see against the incumbents. Descript pioneered edit-by-text, but it's fundamentally a document-first workflow. CapCut is a polished template-and-effects machine aimed at mobile creators. OpusClip is superb at one narrow job — chopping long videos into viral shorts — but it hands you a finished clip, not a workspace.
ChatCut's wager is that creators want the timeline back. The agent does the tedious first pass, but the output is a professional, track-based edit you can iterate on indefinitely — closer to a very fast junior editor than to a one-click generator. That's a meaningfully different value proposition, and it's the reason the tool is drawing attention beyond the usual auto-shorts crowd.
Pricing and the credit system
ChatCut is credit-based. There's a free tier with a small batch of starter credits and no card required, then paid plans built on an adjustable slider:
| Plan | Price | Credits |
|---|---|---|
| Free | $0 | Starter credits, no card required |
| Pro | $25/mo | 100 credits (slider up to 10,000) |
| Team | $100/mo | 400 credits |
Annual billing knocks off roughly 50%. The important nuance: not everything costs credits. Uploading media, automatic transcription, manual timeline editing, playback, previews, and captions built from an existing transcript are all free. Credits are spent mainly on AI generation — new video, images, voiceover — and you see an estimated cost before you run it.
The ChatGPT angle is the sleeper feature
Here's the part that could quietly reshape the math. ChatCut also ships as a plugin inside the ChatGPT / Codex app, and when you drive it that way, editing actions consume your ChatGPT tokens instead of ChatCut credits. If you already pay for ChatGPT, a large chunk of the editing workload effectively becomes free, and you're only buying credits for generative assets. For heavy users, that turns ChatCut from "yet another subscription" into "a front-end for one I already have."
Where it shines — and where it won't
ChatCut is squarely aimed at solo creators, founders, educators, marketers, and social teams turning raw footage into publishable video without learning a professional NLE. On messy interview clips and talking-head content, early users report auto-cuts that respect pacing and, crucially, stay editable afterward.
It is not a finished replacement for a senior editor, and the founders are refreshingly candid about that. Taste, timing, reliability, and performance on long or heavy files are exactly where an AI editing agent is hardest, and they say so out loud. Independent reviews land it around a 7/10 — strong for the use case, not magic. And because generation runs on premium models, credit consumption on generation-heavy projects can climb fast if you aren't watching the estimate.
The Bottom Line
ChatCut is the most convincing answer yet to the "AI edited my video and now I'm stuck" problem. The editable timeline, reusable skills, and XML export mean AI is the start of the edit, not the end of it — and the ChatGPT plugin quietly undercuts the whole category on price for anyone already paying OpenAI. If you make video weekly and dread the timeline grind, spend your free credits on a real clip before you commit.
