News for hackers. Stuff that matters.

A comment on GLM-4.7: Advancing the Coding Capability. Parent comment

Re:GLM-4.7: Advancing the Coding Capability by ImprobableTruth on Monday December 22, 2025 @05:16PM (#46359825)
How is the raw Gemini 3 CoT accessed? Isn't it hidden?
[ Parent | on HN ]
Re:GLM-4.7: Advancing the Coding Capability by Tiberium on Monday December 22, 2025 @05:53PM (#46360165)
There are tricks on the API to get access to the raw Gemini 3 CoT, it's extremely easy compared to getting CoT of GPT-5 (very, very hard).
[ Parent | on HN ]
Re:GLM-4.7: Advancing the Coding Capability by ceroxylon on Monday December 22, 2025 @08:13PM (#46361284)
What are you referring to? I see the 'reasoning' in OpenRouter for GPT-5.2, I was under the impression that is the CoT.
[ Parent | on HN ]
Re:GLM-4.7: Advancing the Coding Capability by Tiberium on Monday December 22, 2025 @08:21PM (#46361333)
Yes, that's exactly what I'm referring to. When you're using the direct Gemini API (AI Studio/Vertex), with specific tricks you can get the raw reasoning/CoT output of the model, not the summary.
[ Parent | on HN ]
Re:GLM-4.7: Advancing the Coding Capability by bwat49 on Tuesday December 23, 2025 @10:05AM (#46365821)
in antigravity gemini sometimes inserts its CoT directly into code comments lol
[ Parent | on HN ]