Artificial Analysis breaks down Intelligence Index v5: Gemini 4 Neon 17% cheaper per task than Carbon
ArtificialAnlys · x · 2026-11-19
Artificial Analysis published the per-evaluation breakdown of its Intelligence Index v5, covering nine independent benchmarks including AA-Briefcase, Omniscience, AutomationBench, Terminal-Bench-Science, FrontierCode-AA, HLE-Diamond, GDPpdf, Terminal-Bench 4.0, and Critpt v2, ranking models on agentic knowledge, hallucination rate, terminal coding, and physics reasoning.
Cost math: Gemini 4 Neon (xhigh) costs $6.44 per Intelligence Index task, 17% less than Gemini 4 Carbon (high, $7.76 at standard pricing) but 1.6x GPT-6.2 Astra (max, $4.00). At high effort Neon matches Carbon's Intelligence score (60) for $3.74 per task — less than half of Carbon's cost.
Related event: Leaked Gemini 4 Neon benchmarks reportedly beat rivals(4 posts)→
More from Models
- Google unveils Gemini 4 Neon: No. 3 on AA Intelligence Index, Pareto frontier pricing — GeminiApp · 2026-11-19
- Gemini 4 Neon benchmark table: tops GraphWalks long-context at 95.2% vs Opus 5.5 and GPT-6.2 Sol — divy93t · 2026-11-19
- Google launches Gemini 4 Neon, debuting at No. 1 on Artificial Analysis Coding Agent Index — Polymarket · 2026-11-19
- Gemini Neon test: it names which "AI SEO courses" are rebranded legacy content and which to skip — gaganghotra_ · 2026-11-19
- Gemini 4 Neon's 39% SWE-sweep questioned: results hinge on the harness — OfirPress · 2026-11-19
- scaling01 mocks benchmark tables: leading every eval doesn't mean the better model — scaling01 · 2026-11-19