Recommended for your Mac trending on Hugging Face · only models that fit
These are the GGUF models trending on Hugging Face right now that fit in your unified memory. Pick one and it drops straight into the approve form on Models you host.
Unified memory
Detecting your device…
Best fits
Fit =
params_b × bytes_per_param + 1.5 GB ≤ 80% of memory · quant defaults to Q4 for the trending view. Same picks from your terminal: python3 scripts/recommend-model.py (a.k.a. umbra recommend). Details in docs/model-recommendations.md.