Jank Fighter II — Claude vs Codex vs Grok before and after Buoy
I built Buoy — debugging and optimize tooling for React Native, Expo, Flutter, Swift, web, and more. We put three coding models through the same prompt on a janky React Native Skia scene rendered with React Native Web in
I built Buoy — debugging and optimize tooling for React Native, Expo, Flutter, Swift, web, and more. We put three coding models through the same prompt on a janky React Native Skia scene rendered with React Native Web in Chrome, then ran Optimize and measured again.
Results (FPS, before → after)
- Claude: 975,800 → 6,123,350 (6.3×)
- Codex: 155,450 → 444,100 (2.9×)
- Grok: 202,100 → 577,350 (2.9×)
Lights held around 55 fps.
Same prompt, same scene, same harness — the gap is what Optimize changed in the app under test.
Full write-up with screenshots and method: AI models before and after Buoy
Questions welcome — especially if you want details on the Skia/Web setup or how we scored the runs.
Originally published by Dev.to AI. Aggregated on AIWithGhost for educational purposes — full credit and traffic to the original publisher.