This product was not featured by Product Hunt yet. It will not be visible on their landing page and won't be ranked (cannot win product of the day regardless of upvotes).
Every other AI video tool works blind. It writes a scene, gets back "success", and never knows whether the frame actually looks right. v2 fixes that: the agent renders real frames and looks at them. It caught a title running off the canvas in our own launch video and patched the one line that caused it. Chat with it in the browser editor, or connect it to Claude, ChatGPT or your IDE over MCP. Free to start, no install.
I'm Emanuele, solo founder. DevMotion is the thing I've been building nights and weekends.
I shipped v1 in April. It made videos from a prompt, and it mostly worked. But it had a flaw I could not stop thinking about: the agent was working blind.
Every tool call came back "success". That only ever meant the data saved — never that the frame actually looked right. Text overflowed its box. Layers stacked on top of each other. And the agent cheerfully reported everything was fine.
I was the eyes, every single time. Which is a strange thing to say about a tool whose whole promise is that you don't have to be.
So v2 gives it eyes. The agent renders real frames and looks at them before it says it's done. When something is off, it goes back and fixes it.
The video on this page is the proof: the agent built it, rendered it, spotted a title running off the canvas, and patched the one line of CSS that caused it. I didn't touch that fix.
You can chat with it in the browser editor, or connect it to Claude or ChatGPT over MCP and make videos from wherever you already work.
What I'd love from you today:
→ Break it. Give it something visually hard and tell me where it still gets it wrong. Those are the reports I learn the most from.
→ If you make videos for a living: is "the AI checks its own work" something you'd actually trust — or does it need to show you the frames it looked at?
Free to start, no install, no credit card. I'm here all day and I answer everything.
No comment highlights available yet. Please check back later!
About DevMotion v2 on Product Hunt
“The AI video agent that looks at what it builds”
DevMotion v2 was submitted on Product Hunt and earned 4 upvotes and 1 comments, placing #24 on the daily leaderboard. Every other AI video tool works blind. It writes a scene, gets back "success", and never knows whether the frame actually looks right. v2 fixes that: the agent renders real frames and looks at them. It caught a title running off the canvas in our own launch video and patched the one line that caused it. Chat with it in the browser editor, or connect it to Claude, ChatGPT or your IDE over MCP. Free to start, no install.
DevMotion v2 was featured in Developer Tools (518.8k followers), Artificial Intelligence (477.9k followers) and Video (2k followers) on Product Hunt. Together, these topics include over 207k products, making this a competitive space to launch in.
Who hunted DevMotion v2?
DevMotion v2 was hunted by Emanuele Pavanello. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how DevMotion v2 stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Emanuele Pavanello
Hey Product Hunt 👋
I'm Emanuele, solo founder. DevMotion is the thing I've been building nights and weekends.
I shipped v1 in April. It made videos from a prompt, and it mostly worked. But it had a flaw I could not stop thinking about: the agent was working blind.
Every tool call came back "success". That only ever meant the data saved — never that the frame actually looked right. Text overflowed its box. Layers stacked on top of each other. And the agent cheerfully reported everything was fine.
I was the eyes, every single time. Which is a strange thing to say about a tool whose whole promise is that you don't have to be.
So v2 gives it eyes. The agent renders real frames and looks at them before it says it's done. When something is off, it goes back and fixes it.
The video on this page is the proof: the agent built it, rendered it, spotted a title running off the canvas, and patched the one line of CSS that caused it. I didn't touch that fix.
You can chat with it in the browser editor, or connect it to Claude or ChatGPT over MCP and make videos from wherever you already work.
What I'd love from you today:
→ Break it. Give it something visually hard and tell me where it still gets it wrong. Those are the reports I learn the most from.
→ If you make videos for a living: is "the AI checks its own work" something you'd actually trust — or does it need to show you the frames it looked at?
Free to start, no install, no credit card. I'm here all day and I answer everything.