RealityBench upheaval: OpenAI's new Terra overperforms, Clef charts new Pareto frontier
arnabdotorg · x · 2026-10-29
- Just 2.5 weeks after RealityBench launched, the top of the physical-world-understanding leaderboard has shifted substantially: Gemini 4 Argon set or matched top performance at 50%–6× cheaper, followed by strong new OpenAI releases 6.1 Sol and 6.1 Terra.
- The author calls Terra's results surprisingly good: "It shouldn't be this good!"
- Adding Clef to the benchmark reveals a new Pareto frontier — decisions are a very cost-effective pathway for physical world understanding (work led by @PerraultAndrew).
More from Models
- Nat Lambert Mocks Teased 'Gemini 4 Carbon' as a Graspy Preemptive Move — natolambert · 2026-10-29
- Unverified 'Gemini 4 Carbon' SOTA claims circulate alongside Google Docs ranking jab — gnukeith · 2026-10-29
- DeepMind's Samuel Albanie: new model's hallucination rate much improved over Argon — SamuelAlbanie · 2026-10-29
- User pushback: US AI models are 'heavily censored' and less useful — gnukeith · 2026-10-29
- Google slammed for slow releases: rivals may outperform Gemini 4 Carbon before it ships — Angaisb_ · 2026-10-29
- Nat Lambert calls out Google: don't announce Gemini scores nobody can try — natolambert · 2026-10-29