Post #3880941
2026-07-17 08:43 UTC
I have a feeling that it won't be "more Claude" that's deciding over whether or not LLMs are "here to stay" but DeepSeek, Kimi, GLM, Mistral etc etc
Any regular, average company who actually WANTS to have LLMs is IMHO better off with the other models - or even chooses to go local-&-mini models, purpose-made which don't cost you an arm and a leg of qualified developer.
Set aside the choice with whom to do business or run it yourself entirely.
It seems German companies at least go this route
Replies (1)
-
@masek@infosec.exchange 2026-07-17 08:46
@sushee@ohai.social Yep, with Turboquant and Qwen 3.6 27B you can do most coding AI stuff inhouse in reasonable quality.