Analysis updated 2026-07-26
Type a prompt to generate a short animated video and download it as an MP4.
Create quick marketing or social media video clips without video editing skills.
Save and manage your generated videos from your profile history.
Install the app on your device as a progressive web app for quick access.
| programmersd21/frame_studio | agentigris/agentfund | alange/tinyclaude | |
|---|---|---|---|
| Stars | 5 | 5 | 5 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | moderate | hard | moderate |
| Complexity | 3/5 | 5/5 | 3/5 |
| Audience | pm founder | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires a Gemini API key and a Supabase project with database tables set up via a SQL script.
Frame Studio is a web app that turns a text prompt into a finished, downloadable video. You type a description of what you want, and Google's Gemini AI writes the code for the animation, compiles it, and renders it as an MP4 file directly in your browser. The project is built with React, Remotion, TypeScript, and the Gemini API, and it runs on Vercel's free tier with Supabase handling accounts and saved files. The process happens in several steps. First, Gemini reads your prompt and plans a video structure. Then it writes the React and Remotion code needed to build that animation. The system checks the code for errors, and if it finds any, Gemini tries to fix them up to three times. Once the code is valid, the server compiles it and sends it to your browser. Your browser then renders the final video using a technology called WebCodecs, which lets it encode video without sending the work to a remote server. You get an MP4 file immediately, with no waiting in a queue. To run Frame Studio yourself, you need Node.js, the pnpm package manager, a Gemini API key, and a Supabase project. After installing dependencies, you copy a settings file and add your keys for Gemini and Supabase. You also run a SQL script in Supabase to set up the database tables for storing video data. You can start the app locally with a single command, and you can optionally enter your Gemini key directly in the web interface instead of the settings file. The app includes user accounts with email and password sign-in, magic links, password resets, and profile settings where you can upload an avatar. Saved videos appear in your profile history, and you can preview or delete them there. The interface is designed to look polished, with smooth animations and a layout that adapts to mobile and desktop screens. It also works as a progressive web app, meaning you can install it on your device. Video rendering relies on the WebCodecs API, so it only works in newer browsers: Chrome 94 and above, Edge 94 and above, or Firefox 130 and above. Videos are rendered at 1920 by 1080 resolution and 30 frames per second. The project is open source under the MIT license.
Type a text prompt and Google's Gemini AI writes the animation code, renders it in your browser, and gives you a downloadable MP4 video file, no waiting in line.
Mainly TypeScript. The stack also includes React, Remotion, TypeScript.
Use freely for any purpose, including commercial use, as long as you keep the copyright notice.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly pm founder.
This repo across BitVibe Labs
Verify against the repo before relying on details.