About running AI LLM models locally and privately:If you plan to get this and run AI, know that up to 32 GB leaves you hardly any space for SOTA LLM models:
32 GB RAM - 8 GB for the OS itself = 24 GB free for fitting an AI LLM model.
The current SOTA AI LLM models, for their size, are:
huggingface.co/Qwen/Qwen3.6-27B (dense architecture) (Downloads last month: 5.0 million)
and
huggingface.co/Qwen/Qwen3.6-35B-A3B (MoE architecture) (Downloads last month: 5.6 million).
A popular, bang for the buck, quant is huggingface.co/unsloth/Qwen3.6-35B-A3B-MTP-GGUF UD-Q4_K_XL (22.9 GB) (good for 80k-120k of context). Additional 2 GB for 32k and 8 GB for 128k (linear)[1] of context are required. This means that the quant barely fits and there's pretty much almost or no space left for any context, see e.g. this quote:
Quote from: reddit.com/r/LocalLLaMA/comments/1sq94qx/is_anyone_getting_real_coding_work_done_with.. I've come to the conclusion that (1) 32768 is the biggest context I can get away with in an adequately smart model, and (2) it just ain't enough.
In a dense 27B model, all 27B parameters per token are used vs in the 35B-A3B, where only 3B active are used. Hence, for the same size, a dense model performs much better[2]. The 35B-A3B has the advantage of running faster. You can also see it as a speed (MoE model) vs size (dense) trade-off. A MoE model is more suited for RAM speeds, but since RAM prices have increased by like 4 times, the advantage of a MoE model went away and we are back to running dense models off GPU's much faster VRAM. There need to be more dense models for this reason.
huggingface.co/unsloth/Qwen3.6-27B-MTP-GGUF UD-Q4_K_XL quant (17.9 GB) runs slower, but at least it does somewhat fit with some context.
What would be needed are additional 8 GB VRAM dGPU (then it would be basically a gaming laptop) or a total of 48 GB system RAM.
Unfortunately, this laptop only goes up to 32 GB RAM. A 48 GB RAM option is required to get rid of the limitation.
[1] reddit.com/r/LocalLLaMA/comments/1tvluaj/how_much_vram_needed_for_qwen_36_27b_q8_with_262k
[2] See 27B dense vs 122B-A10B (MoE) evaluation: huggingface.co/Qwen/Qwen3.5-122B-A10B#benchmark-results. The 122B required 4.5 times more memory, but they score about the same.
Deutsch
English
Español
Français
Italiano
Nederlands
Polski
Português
Русский
Türkçe
Svenska
Chinese
Magyar