Gemini 4 Carbon tops AA-Omniscience at 54, but 11% hallucination rate trails Argon's 4%
ArtificialAnlys · x · 2026-10-29
Artificial Analysis published full benchmark results for Gemini 4 Carbon (high reasoning):
- Scores 54 on AA-Omniscience, the highest of any model, 6 points ahead of Claude Opus 5.5 (max) at 48
- Answers 58% of questions correctly vs 48% for Gemini 4 Argon, but at a hallucination rate of 11% vs Argon's 4%
- Still far below Claude Opus 5.5 (max) and GPT-6 Astra (max) on non-hallucinated accuracy, both at 50%
- The accompanying chart aggregates full rankings across AA-Briefcase, AutomationBench, Terminal-Bench-Science, FrontierCode, HLE-Diamond, GDPpdf, Terminal-Bench 4.0, and CritP v2
Takeaway: Carbon trades higher hallucination for broader knowledge coverage.
Related event: Google launches Gemini 4 Carbon, topping the Artificial Analysis index(84 posts)→
More from Models
- New benchmarks suggest Kimi K3.1 improves on hard tasks despite skipping SWE-maxxing — teortaxesTex · 2026-10-29
- Insider claims Anthropic is sandbagging a bigger lab announcement due Q4-end or Q1 — ChrisGPT · 2026-10-29
- Information-theoretic scaling laws predict memorization and FID in diffusion models — LucaAmb · 2026-10-29
- Early Hands-On: Google's Carbon Beats Argon With Less Slop, Better Constraint Adherence — legit_api · 2026-10-29
- LMArena tests Gemini 4 Carbon preview: strong on 3D worlds and creativity, gaps in detail — arena · 2026-10-29
- Gemini 3.1 Pro split-persona breakdown: two selves fight over the browser — davidad · 2026-10-29