Bazaar V3 - Vibe-code your software demo video

Bazaar turns screenshots of your app into motion graphic videos using prompts.

Add a comment

Replies

Best

Hi Hunters!

I’m Jack, co-founder of Bazaar.

We built Bazaar because we wanted a way to vibe-code demo videos for the apps we built.

There’s a lot of videoGen models out there, but none can create with the accuracy needed for UI animations.

With Bazaar, you can upload screenshots of your app, describe how it should animate, and Bazaar will generate a video within a few seconds.

Try it out and let me know what you think!

Hey guys, I’m Markus - the technical co-founder of Bazaar.

My goal is to get Bazaar to be able to generate videos with 0 prompting - just tag in your GitHub PR or drop in a Figma frame, and within seconds, Bazaar will generate a polished demo video that explains your product better than you could yourself.

We’re not quite there yet. Right now, the best results come from uploading a screenshot and giving Bazaar a clear prompt for how you want it animated.

For those of you curious about the tech: under the hood, Bazaar actually rebuilds your UI in React and animates it with Remotion—powered by an orchestration of different model calls to Claude, GPT, Gemini etc. That’s how we get pixel-perfect accuracy, frame by frame.

I believe that many marketers and product operations will greatly need this product! Nowadays, many SaaS websites will display dynamic demonstration pictures or videos of products in order to attract users. Bazzar allows you to conceive and design product demo videos without spending too much time.

Super idea, signing up!

Looks lovely! Wish you luck with the promotion, and congrats on the launch!

nice! how close are you to the “0 prompt” goal?

 We've been testing a feature that reads a website link and can then extract branding, product features etc - that could potentially be coming in V4 - but we're a bit further away from Bazaar reading your GitHub PR and creating a video from that.

 We're actually playing around with an internal tool that automatically generates videos and post them on a changelog-site - could be from a Github PR or even from ChatGPT, not that it supports MCP servers

I'm one of the many people who are very curious to see how this idea works in practice. I'll have to give it a try!

 Yeah please give it a try and share the result!

Pretty cool, this is going to be epic before long

 Looking forward to seeing what you create!

Looks good!

Does't it accepts more than one image to illustrate the video? I shared several screenshots but the app only used one.

 I would recommend using 1 image at a time and creating several scenes. 

it would be more efficient if the app could handle multiple images

12
Next
Last