Back Original

llm-gemini 0.38.1

Gemini 4 Carbon only supports low, medium and high reasoning levels - but Gemini 4 Hydrogen and Gemini 4 Neon both have an extra xhigh level.

I missed that when I added those models to the llm-gemini plugin, which is now fixed in this dot release.

The xhigh setting for Hydrogen makes a significant difference. I updated my Gemini 4 comparison grid - note how Hydrogen at high used 5,311 output tokens and cost 3.99 cents, but the same model at xhigh used 31,477 output tokens, cost 23.61 cents, and produced a pelican that's competitive with Gemini 4 Neon:

Screenshot of a comparison table of illustrations of a pelican riding a bicycle, generated by three models at two thinking levels. Column headers: Thinking, gemini-4-carbon, gemini-4-neon, gemini-4-hydrogen. Row XHIGH: gemini-4-carbon cell is empty and reads "Unsupported". gemini-4-neon shows a detailed white pelican with a large orange beak riding a red cruiser bicycle along a beachside road, a fish in the wicker basket on the handlebars, with sea, a sailboat, clouds and sun behind, captioned "40 input tokens, 26,546 output = 39.83 cents · 3m 47s". gemini-4-hydrogen shows a similar pelican on a teal bicycle on a sandy beach with a blue fish in the basket, captioned "40 input tokens, 31,477 output = 23.61 cents · 2m 36s". Row HIGH: gemini-4-carbon shows a pelican wearing a sailor hat riding a red cruiser bicycle with a fish in the basket beside the sea, captioned "40 input tokens, 20,305 output = 20.31 cents · 4m 11s". gemini-4-neon shows a pelican on a simpler red bicycle on a dark asphalt road with yellow dashed lines and the sea behind, captioned "40 input tokens, 9,691 output = 14.55 cents · 1m 11s". gemini-4-hydrogen shows a plainer, flatter pelican on a red bicycle over a green strip of ground with no basket or sea, captioned "40 input tokens, 5,311 output = 3.99 cents · 19s".