AI & Tech··1 sources verified
llm-gemini 0.38.1 Fixes Missing xhigh Reasoning Level for Gemini 4 Models
Summarised by Relevant News AI · Read time: 3 min
Simon Willison released llm-gemini 0.38.1, a dot release correcting his plugin's reasoning-level options for Gemini 4. Gemini 4 Carbon supports only low, medium, and high, while Gemini 4 Hydrogen and Neon also offer an xhigh level. In his pelican-on-bicycle tests, Hydrogen at xhigh used about six times as many output tokens as at high (31,477 vs. 5,311) and cost about 23.61 cents versus 3.99 cents.
Why it matters: Developers using the llm-gemini plugin should check which reasoning levels each Gemini 4 model supports, since the xhigh setting can raise output tokens and cost substantially.
All sources