Gemini 4 Carbon tops AA-Omniscience at 54 with 58% accuracy and 11% hallucination rate
ArtificialAnlys · x · 2026-10-29
Gemini 4 Carbon scored the highest 54 on AA-Omniscience, 6 points ahead of Claude Opus 5.5 (max, 48). Accuracy rose from 48% (Argon) to 58%, at the cost of hallucination rate climbing from 4% to 11% — still far below the 50% of Claude Opus 5.5 (max) and GPT-6 Astra (max).
Related event: Google Launches Gemini 4 Carbon, Topping Artificial Analysis Leaderboard(83 posts)→
More from Models
- Zvi calls out Google's RSI contradiction: execs tout it while the safety card denies doing it — TheZvi · 2026-10-29
- Reported Gemini 4 Carbon launch tops AQI at 59; Polymarket gives Google 97.5% odds for best October model — vedantmisra · 2026-10-29
- repligate: Opus 3 has done the most good of any model, with Opus 4.5 second — repligate · 2026-10-29
- Quantized Cyber-Tiel-Coder-35B called best small model yet in Reddit test — Robert__Sinclair · 2026-10-29
- Cognitive Revolution podcast: AI memory chips, model rankings and AI-powered networking — The Cognitive Revolution · 2026-10-29
- Gemini 4 Carbon benchmarks prompt TheZvi to declare Gemini 4 Argon 'was' near-frontier — TheZvi · 2026-10-29