3 ms·
Where can a user reasonably host this in an affordable way to access the local LLM revolution?
by aliljet 5mo ago
Where can a user reasonably host this in an affordable way to access the local LLM revolution?
- truetotosse 5mo agoThis one is not local
- plagiarist 5mo agoI think their Max models are far bigger than fits on consumer hardware. People are typically using Apple, AMD Halo, or dGPUs if/when they do smaller versions. Those are all varying degrees of "affordable."
- julianlam 5mo agoTry llama.cpp and Qwen3.6-35B-A3B Good balance of intelligence and speed.
- satvikpendem 5mo agoUnsloth Studio with its MTP support: https://unsloth.ai/docs/models/qwen3.6#mtp-guide https://unsloth.ai/docs/models/qwen3.6#mtp-guide