Recent post re: AI as utility
Myself, I’m a fan of local LLM / self hosted ML… but if you ever needed a clarion call that a hard pivot is coming (soon) for online/ cloud based AI…Altman et al are making some concerning mouth noises (to say nothing of broader concerns with OAI, Anthropic etc).
Right now, I’m sketching out a plan where my Raspberry Pi (always on, 2-3w) uses a magic packet to wake up my modest AI server (Lenovo P330 with Tesla P4) if/when needed (Qwen 3.6-35B-A3B); no point in chugging down 80-100w, 24/7 for no good reason.
If the trend continues the direction it appears to be (increasing costs, environmental impacts etc) then I’d feel a lot better hosting my own as port of first call and replacing simpler tasks with more traditional programs. YMMV.
Does anyone have a recommendation for a local model that can run well on a 5070 12GB? It pretty much would only get used for help with homelabbing and simple scripts.
Depends on how much CPU RAM you have, and how fast it is.
As others said, Qwen 35B at the very least. But you can get better models with more CPU RAM.
Ive got 32GB DDR5 6000mhz
Probably Qwen 35B then. ~9GB free VRAM + (let’s say) ~16GB of free CPU RAM is a good size for that, and squeezing bigger models in would be hard unless it’s a headless linux server.
There’s an argument to be had regarding a MoE versus a small dense model. I guess it depends on what exactly you need doing with it. I would be tempted to run a smaller dense model (like a Qwen 3-14B or a Qwen 3.5 9B) as at a reasonable quant, it might fit mostly or entirely on the GPU, thereby giving you excellent speeds.
PS: I’m actually in the process of designing an expert system (not a LLM) for pretty much the task you described. The intention is that you would still interact with it like a large language model, but the actual brains underneath it would be something more traditional.
MoEs can be very fast with hybrid inference. I run Xiaomi Mimo 2.5 (a 310B model, 116GB weights) on my single 3090 + 7800 CPU, and it outputs faster than I can read it.
It’s also easier to fit long context, if you need that.
It’s best to use the ik_llama.cpp fork for that, though. It gives a huge boost to hybrid MoE speeds.
Qwen 3.6-35B-A3B (which OP mentioned) would work great as long as you have some system RAM to offload it.