Rumored Claude Fable 5.5 beats Opus 5.5 and Gemini 4 Argon on three.js giraffe test
minchoi · x · 2026-11-03
minchoi ran the same three.js prompt (a running giraffe) through rumored "Claude Fable 5.5" (via routing), Opus 5.5, and Gemini 4 Argon. Fable 5.5 was the only output whose gait, legs and grass moved like a real animal, clearly beating Opus 5.5 on this test. He also shared a three.js car scene where body, paint, lighting and camera all run as JavaScript in the browser — hard to believe it's not a render. Note: Fable 5.5 is an unannounced model, so findings are third-party tests, not official benchmarks.
Related event: Fable 5.5 vs Opus 5.5 and GPT-6.2 in three.js showdown(2 posts)→
More from coding & agent
- EAT Evolves Agent Teams to Hit SOTA on Kernel Optimization, Ranks #4 in Open Kernel Challenge — berkeley_ai · 2026-11-03
- Sol 6.1 still best cost-per-intelligence worker; Fable 5.5 more well-rounded — haider1 · 2026-11-03
- Zed founder: visuals are the biggest barrier to AI game design, no working pipeline yet — zeeg · 2026-11-03
- Dividing attention isn't dividing context: Google's Deep Review multi-agent design — krishnan · 2026-11-03
- LangChain: Agentic browser requests jumped ~5x in September alone — LangChain · 2026-11-03
- OpenCode rebuilt to run inside browser DevTools, betting the next billion-dollar apps live there — nicolascraske · 2026-11-03