Gemini 4 Neon debuts at #2 on Vals Index at less than half Claude's cost
demishassabis · x · 2026-11-19
Google DeepMind CEO Demis Hassabis retweeted Vals AI's latest benchmark: Gemini 4 Neon debuts at #2 on the Vals Index with 70.87% accuracy, just 0.4 points behind Claude Fable 5.2 (71.31%) at $7.40 per task — less than half of Fable's $17.94 — and with lower latency (31m25s vs 43m42s).
It also takes #1 on the Finance and Legal Indexes, plus Tax Agent and Legal Research. The top five also includes Claude Sonnet 5.5 (70.85%), Gemini 4 Carbon (70.45%) and Claude Opus 5.5 (70.14%); GPT-6.2 Astra sits 8th at 67.66%.
More from Models
- Gemini audits 30k lines written by Opus 5.5, finds 29 real bugs with zero false positives — neurocy · 2026-11-19
- Top 10% OpenAI researchers reportedly burn $25,000 of tokens a day — ivan_bezdomny · 2026-11-19
- Claude Max for $29 instead of $200: reseller gray market sparks debate — mehdiweb · 2026-11-19
- Chris Albo backs "super intelligence" as a term that separates math-discovering systems from a random forest — chrisalbon · 2026-11-19
- Rumor (started by the poster): SSI's first model will be named 'superintelligence' — daniel_mac8 · 2026-11-19
- Leaked Gemini 4 lineup uses element names: Hydrogen, Argon, Carbon and new Neon tiers explained — dreamwieber · 2026-11-19