Comment

As far as I understand, the only way to get anything resembling usable output for coding is with massive, expensive, labouriously hand-tuned models, not local ones.

source

Sort:hotnew top

locuester@lemmy.zip ⁨6⁩ ⁨days⁩ ago
Qwen 3.6 27B dense is really good. Very usable coding output

source
irelephant@lemmy.dbzer0.com ⁨6⁩ ⁨days⁩ ago
Deepseek is pretty good the few times I tried it.

source
Jestzer@lemmy.world ⁨6⁩ ⁨days⁩ ago
^^^ This. Tragically, locally run LLMs don’t even hold a candle to “good” cloud-based LLMs like Claude Code.

source
Kkk2237pl@lemmy.world ⁨6⁩ ⁨days⁩ ago
I see that qwen 3.5 has pretty good performance and can be run on macbook with 64GB ram

source
- SmoothLiquidation@lemmy.world ⁨6⁩ ⁨days⁩ ago
  I have played with qwen3-coder:30b for my hobby stuff running on my M5 max MacBook and it does alright. It is fast enough and I used ollama tools to let it request files. I haven’t used anything like Claude code to compare it to though, only a bit of the ChatGPT free tier stuff.
  
  source
- Penta@lemmy.world ⁨6⁩ ⁨days⁩ ago
  Qwen 3.6 is even better
  
  source