Comment Comment by dirkg > For 16GB laptops, Qwen 3.5 9B is the undisputed champ.you can run qwen 3.6 35BA3B on a 12-16GB vram gpu and ot works pretty well.https://www.youtube.com/watch?v=8F_5pdcD3HY&t=1seven the 27B in some quants can fit.https://www.reddit.com/r/LocalLLaMA/comments/1tkmgwj/qwen27b...qwen IMO is far better for coding, esp agentic coding when combined with something like Pi, it comes probably close enough to Sonnet for a lot of use cases.Gemma family is better for almost all other tasks you'd use a local llm for. parent dirkg 5 months ago
you can run qwen 3.6 35BA3B on a 12-16GB vram gpu and ot works pretty well.
https://www.youtube.com/watch?v=8F_5pdcD3HY&t=1s
even the 27B in some quants can fit.
https://www.reddit.com/r/LocalLLaMA/comments/1tkmgwj/qwen27b...
qwen IMO is far better for coding, esp agentic coding when combined with something like Pi, it comes probably close enough to Sonnet for a lot of use cases.
Gemma family is better for almost all other tasks you'd use a local llm for.