Leaked-looking LiveBench screenshot puts Gemini 4 Neon first at 87.4
Dense_Primary629 · reddit · 2026-11-22
A Reddit user posted a screenshot purportedly from livebench.ai (unverified, odd model names, authenticity doubtful): Gemini 4 Neon ranks first overall at 87.4, ahead of Claude Fable 5.2 (xhigh) and GPT-6.2 Astra (both 84.9). GPT-6.2 Astra leads coding (91.9), Claude Fable 5.2 leads design (85.1). The unfamiliar naming suggests the chart may be fabricated or placeholder data — treat with caution.
More from Models
- Fine-tune gotcha: HF export silently drops Qwen3's MTP head, slowing decode — AIFlow_ML · 2026-11-22
- Designer KOL: Google's model pace and quality suggest it will soon 'mog them all' — AIandDesign · 2026-11-22
- Models play a literary 'listening at doors' game — Opus 4.5 gets recognized by its timbre — repligate · 2026-11-22
- Asking Claude to 'understand the entire mathematical platonic realm, make no mistakes' — burny_tech · 2026-11-22
- Twitter talks about Google Neon, but Google AI's landing page pushes Carbon — annetgriffin · 2026-11-22
- Viral claim: Claude produced a 250,000-page Riemann hypothesis proof — tinyfool · 2026-11-22