text model · Nemotron 3 · macOS
Can I run Nemotron 3 Super 120B-A12B on Apple M5 Max (128GB)?
Yes. Nemotron 3 Super 120B-A12B runs on Apple M5 Max (128GB) at Q4_K_M (~80.3 GB of ~96 GB usable).
Runs at Q4_K_M using ~80.3 GB of ~96 GB usable.
That figure is at a 4k context and moves about ±15% as context length changes. Apple M5 Max (128GB) leaves ~15.7 GB of headroom.
- Q4_K_M needed
- ~80.3 GB
- Usable on device
- ~96 GB
- Device memory
- 128 GB
- Best quant
- Q4_K_M
Which quant fits
Run it
Pick your tool. All 3 load the same Q4_K_M weights.
ollama run nemotron-3-super:120b-a12b llama-cli -hf unsloth/NVIDIA-Nemotron-3-Super-120B-A12B-GGUF:Q4_K_M lms get unsloth/NVIDIA-Nemotron-3-Super-120B-A12B-GGUF - Parameters
- 120B (MoE, 12B active)
- Q4_K_M size
- 76.87 GB
- Q8_0 size
- 119.65 GB
- Context
- 256k
- Ollama tag
- nemotron-3-super:120b-a12b
- Memory
- 128 GB unified
- Usable for weights
- ~96 GB
- Best runtime
- MLX direct / Ollama (MLX backend)
You could also run
Run Nemotron 3 Super 120B-A12B on other hardware
FAQ
Can Apple M5 Max (128GB) run Nemotron 3 Super 120B-A12B?
Yes. Nemotron 3 Super 120B-A12B runs on Apple M5 Max (128GB) at Q4_K_M (~80.3 GB of ~96 GB usable).
How much memory does Nemotron 3 Super 120B-A12B need?
Apple M5 Max (128GB) has room to spare. At Q4_K_M the weights are ~76.87 GB; with KV cache and runtime overhead, budget ~80.3 GB at a 4k context. It is a Mixture-of-Experts model (120B total / 12B active), so all experts must stay in memory; memory tracks total params, not active params.
What is the best tool to run Nemotron 3 Super 120B-A12B on macOS?
LM Studio for a simple setup; mlx-lm for the most speed. vLLM is NOT a Mac tool, it is a CUDA/Linux serving engine. Unified memory is not a fixed VRAM slice; ~70% is usable for weights.
Embed this
A live badge for your model card or README, updated as the data is. [](https://localmodel.run/can-i-run/nemotron-3-super-120b-a12b/apple-m5-max-128gb) Sources
Weights are measured from GGUF files; KV cache and overhead are computed, so totals can vary ~15% with context and runtime. Any tok/s is a bandwidth estimate. See methodology.