Simon Willison: Gemini 4 Hydrogen 'xhigh' thinking costs 6x more but is far better
film_girl · x · 2026-11-22
Simon Willison shares a Gemini 4 tip: the gap between "high" and "xhigh" thinking effort on Gemini 4 Hydrogen is very significant. In his pelican-on-a-bicycle test, high used 5,311 output tokens at 3.99 cents (19s), while xhigh used 31,477 tokens at 23.61 cents (2m36s) — roughly 6x the cost for visibly richer output. Worth it for quality-critical tasks, overkill for routine ones.
More from Models
- Tesla FSD V15 rolls out with 25x better safety handling and Reality View — Scobleizer · 2026-11-22
- Uncensored 26.4B Inkling 2 Mini runs locally on a 12GB laptop, keeps 98% MMLU — udmrzn · 2026-11-22
- Tesla starts rolling out FSD v15 with 25x better safety-critical handling — ArnoCandel · 2026-11-22
- Joke tweet suggests Grok 4.5 vs 4.7 teams 'need to be studied' — zdhnarsil · 2026-11-22
- Researcher argues LLM companies have no moat as Gemini 4 Neon kills the data flywheel thesis — adam_karvonen · 2026-11-22
- Anthropic releases FlashProve with Lean formalization, proving 400B MoE inference on one H100 — jiqizhixin · 2026-11-22