Google launches Gemini 4 Neon: frontier intelligence at lower cost, tops AutoBaxBench at 76%
melvinjohnsonp · x · 2026-11-19
Google officially introduces Gemini 4 Neon, pushing frontier-level intelligence into a faster, more cost-efficient model for complex daily work, with enthusiastic internal usage reports. On AutoBaxBench, Gemini 4 Neon scores 76.0%, ahead of Gemini 4 Carbon (74.0%), Claude Fable 5.2 and Claude Opus 5.5 (68.0% each), GPT-6 Astra (57.0%) and Muse Wonder (49.0%).
Available today via:
- The Gemini API in Google AI Studio and Gemini for business;
- The Gemini app for Google AI Pro and Ultra subscribers;
- Google Antigravity for paid Pro/Ultra plans.
Related event: Google Launches Gemini 4 Neon, Topping Coding Agent Rankings(74 posts)→
More from Models
- Rare sight: broad, cross-constituency public praise for Gemini models — matt_slotnick · 2026-11-19
- Seedream 5.0 Pro accused of charging credits while showing 'Completed' with no output — JohnTheWorldfucker · 2026-11-19
- Gemini 4 Neon takes #1 on Coding Agent Index at $3/$15, ending Anthropic's top-3 sweep — olcan · 2026-11-19
- OpenAI taps DeepBinary to eval GPT-6.2 Astra and Sol on Windows kernel vuln discovery — LingmingZhang · 2026-11-19
- Users demand return of ChatGPT's Orion voice after update makes new voices uncanny — FluffyAras · 2026-11-19
- TensorFold 1.2.10: 57% faster Qwen3.8-27B decode on GB10, plus new Metal models — AIFlow_ML · 2026-11-19