Gemini 4 scores 53 on AA intelligence index at a third of Opus 5.5's cost
Sea_Physics401 · reddit · 2026-10-01
According to Artificial Analysis results shared on Reddit, Google's Gemini 4 scores 53 on the intelligence index—a strong result—while costing only about one third of Opus 5.5, making it notably cost-efficient. Details are sparse; see the original leaderboard.
More from Models
- Polymarket claims GPT-6 Astra cracked a 217-year-old Napoleonic military cipher — Polymarket · 2026-10-01
- As model releases pile up, auto-routing between models may become standard for consumers — AnneliesGamble · 2026-10-01
- "No more pacing": developer flags quiet end to AI usage throttling — vivekhaldar · 2026-10-01
- Google models ace benchmarks but feel mid in use — will Gemini 4 Argon differ? — VraserX · 2026-10-01
- Jev reranker cuts cost to 75% while lifting recruiting search matches by 30.64% — hardimanjames · 2026-10-01
- minchoi's monthly roundup: 10+ model releases in one month from GPT-6 to Grok 4.7 — minchoi · 2026-10-01