Google quietly rewrote Gemini 4 Neon docs, dropping "Flash-tier cost" disclosure
lyraxana · x · 2026-11-19
Users spotted that Google silently edited Gemini 4 Neon's model card: the pre-release description said Neon "builds on Gemini 4 Hydrogen" (RL on top of it) with "Flash-tier speed and cost." Both mentions were removed after launch, replaced with "frontier-level coding, scientific reasoning, and multi-step agentic execution" positioning between Hydrogen and Carbon. The change likely hides that Neon is effectively a Flash-class model, avoiding complaints about its price. Before/after screenshots document the edit.
More from Models
- Free Gemini users see model picker shrink to Auto-only within a single month — python_buddy · 2026-11-19
- Gemini 4 Neon High matches xHigh scores at 42% lower cost — adonis_singh · 2026-11-19
- Google shipped Argon, Carbon, Hydrogen and Neon in 5 weeks of Gemini 4 iteration — Afinetheorem · 2026-11-19
- GPT 6.2 Astra shows surprisingly strong zero-shot crystal generation, a year after the task looked hopeless — xennygrimmato_ · 2026-11-19
- Debate Benchmark adds 8 models: Qwen 4 Max jumps 144 points, Claude Fable 5.2 leads at 1688 — zero0_one1 · 2026-11-19
- LLMs zero-shot beat fine-tuned world action models on 7 of 10 robot tasks — paigeinsf · 2026-11-19