Gemini 4 Carbon tops Artificial Analysis Intelligence Index at 59, second only to Opus 5.5 on coding agents
ArtificialAnlys · x · 2026-10-29
Artificial Analysis published first benchmarks for Gemini 4 Carbon: it takes #1 on Intelligence Index v5 with 59, and ranks second on the Coding Agent Index behind Claude Opus 5.5.
- Leads 4 of 9 index evals (HLE-Diamond, Terminal-Bench-Science, AA-Omniscience, AutomationBench-AA)
- Coding agents: +6 over Gemini 4 Argon; #2 on Terminal-Bench 4.0 (64%) and FrontierCode v1.1 (51%)
- Agentic knowledge work improved but not frontier: AA-Briefcase 1609 Elo, behind Grok 4.7, Muse Wonder, Kimi K3.1 Pro
- Cost per task $7.76, 23% cheaper than Claude Opus 5.5 (max); same $4/$20 pricing as Argon but +9 points
- AA-Omniscience 54 (#1), 58% accuracy, 11% hallucination rate (second-lowest)
- 1M context; limited rollout with 50% launch discount
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