Video, founded by Alessio Battagliero, lets users direct edits in plain language, shows its plan before each change and keeps every step reversible ahead of its Product Hunt launch on October 17, 2026
For decades, video editing software has followed a familiar layout: a timeline along the bottom of the screen, a preview window above it, and tracks, keyframes and effects panels in between. Learning to edit has largely meant learning that interface.
GPT-Video, a startup founded by Italian entrepreneur Alessio Battagliero, starts from a different premise. Its users direct the edit by chat. The company calls the approach “vibe editing,” a nod to “vibe coding,” the practice of describing software in plain language and letting an AI model write the code.
In a conventional editor, every change is a manual operation. Removing pauses means finding each one on the timeline and cutting it out. Adding captions means typing or importing them, then adjusting their timing line by line.
In GPT-Video, the same work starts with a sentence. A user might type “remove the silences and add captions,” “caption it in my brand style” or “split-screen this with gameplay.” The editing assistant interprets the request and carries it out through the product’s editing pipeline.
The timeline has not disappeared. According to the company, it still exists underneath. The difference is that the user no longer has to operate it directly.
The Plan Comes First
Handing an edit to software raises an obvious question: what exactly will it do? GPT-Video answers that before anything changes. When a user makes a request, the tool presents its plan for the edit and applies it only afterward.
The same principle covers the choices the system makes on its own. When it suggests segments from a recording, each suggestion carries the reasons it was selected, so the user can evaluate it with context rather than take it on faith.
“Vibe editing means you describe the edit in plain language and the tool does the cutting,” said Alessio Battagliero, founder of GPT-Video. “You can see why each clip was picked, and you can undo any step.”
Undo, One Step at a Time
Every change GPT-Video makes is recorded as a named, separate step, and any of them can be reversed in one click. That matters because conversation is iterative. A first instruction rarely produces the final result, and people tend to refine their requests as they see output.
A session might move from “make the captions bigger” to “go back to the previous style” to “cut the first ten seconds.” When a result misses, the user says so and the tool adjusts it. The agent does not publish or overwrite anything on its own, and finished exports remain files the user controls.
“A conversation is how people already give feedback to an editor,” Battagliero said. “GPT-Video shows its plan before every edit, so you always know what is about to change.”
The structure keeps the user in the role of director. The software performs the mechanical operations, while judgment about what works stays with the person.
Beyond the Cut
The editor also covers tasks that often require separate tools. On every plan, users can remove silences and filler words, build split-screen layouts, generate AI voiceovers in English and Italian, and strip vocals or background music from audio. Fake text message videos and narrated Reddit-style stories are available as native formats.
One of its main uses is turning long recordings into captioned vertical clips for TikTok, YouTube Shorts and Instagram Reels. That is one workflow among several that the chat interface can drive.
Connections and Plans
On the Creator and Studio plans, GPT-Video can connect to outside tools and automations, including AI assistants that support MCP, the Model Context Protocol, an open standard for linking assistants to external software. That allows an assistant or an automation to work on videos without the user opening the app.
Pricing starts at $19 per month for Starter. Creator costs $49 per month and adds saved brand styles, 4K export and room for a small team. Studio costs $199 per month. Each message sent in the chat counts as one request, with daily allowances that rise by plan.
A Test for Chat-Based Editing
GPT-Video is set to launch on Product Hunt on October 17, 2026. The debut will test whether a chat window can serve as the main controls of a video editor. The company’s case rests on transparency: show the plan, explain the choice and make every step reversible.
