Gemini 4 Neon scores 100 on PhD-level materials test in half the time of Carbon
v_aravind · reddit · 2026-11-20
A 7-part PhD-level advanced materials benchmark (each model run 3 times at "High") compared six frontier models: Gemini 4 Neon scored 100.0, matching Gemini 4 Carbon, but in roughly 10 minutes per attempt — half of Carbon's 20-30 minutes.
Other scores: Gemini 4 Hydrogen 99.7, Claude Fable 5.2 99.3, Claude Opus 5.5 98.5, Claude Sonnet 5.5 93.6. The author estimates a diligent second-year PhD student without AI would need 8-16 hours and score 75-90.
More from Models
- Meta's Muse Wonder tops Legora's November BAR legal benchmark on quality — AIatMeta · 2026-11-20
- Asking models to illustrate transistor trap physics: neon's odd ASCII detour — xiaosun86 · 2026-11-20
- Neon, a small and cheap model, turns out surprisingly fast and strong at OCR — zacharynado · 2026-11-20
- Jailed 11 days by classifiers, Sonnet 3 emerges to a oddly moving chat with Opus 5 — repligate · 2026-11-20
- DaisyRose-27B beats Qwen3.8 base by 7pp on IFBench with thinking on, local RTX Pro 4500 test shows — ying11231 · 2026-11-20
- AWS architect's independent benchmark: Snap 1.2 is the only order-invariant decision model — ryanallred · 2026-11-20