run it locally

Run Command R 35B on Apple M3 Max 128GB

35B parameters on a 128 GB (unified memory) card. It fits at FP16 (full) — here's the VRAM math at every quantization, the speed to expect, and the quality you trade.

verdict · Command R 35B on Apple M3 Max 128GB
Fits at FP16 (full)
needs 84.0 GB at FP16 (full) · 92.2 GB usable · ~5 tokens/sec
VRAM needed and fit for Command R 35B on Apple M3 Max 128GB by quantization.
quantization vram needed fits 128gb? tokens/sec quality
FP16 (full)recommended 84.0 GB ✓ yes ~5 Reference quality — no quantization loss.
Q8_0 44.5 GB ✓ yes ~9 Near-lossless; rarely worth the extra space over Q6.
Q6_K 34.4 GB ✓ yes ~12 Virtually indistinguishable from full precision.
Q5_K_M 29.0 GB ✓ yes ~14 Minor loss; an excellent quality-vs-size balance.
Q4_K_M 23.5 GB ✓ yes ~17 Small but measurable loss; the popular default.
Q3_K_M 18.1 GB ✓ yes ~23 Noticeable degradation; only when you're tight on VRAM.

Weights = params × bytes/weight, +20% for KV cache & runtime; usable VRAM is 72% (Apple unified memory) of nameplate. Tokens/sec is a bandwidth ceiling (400 GB/s) — real throughput is lower with long context. Try other combinations →

Running Command R 35B on a Apple M3 Max 128GB

At FP16 (full), Command R 35B's weights take about 84 GB, inside the Apple M3 Max 128GB's 92 GB of usable memory. Reference quality — no quantization loss. Generation speed is bound by memory bandwidth — the GPU reads the whole model once per token — so expect on the order of 5 tokens/sec before context overhead.

Quantization is the lever

Each step down in precision shrinks the model: FP16 needs about 84 GB, Q4_K_M about 24 GB — a 72% reduction for a small, usually acceptable quality cost. Q5_K_M and Q6_K are near-lossless if you have the headroom; drop to Q3 only when you're genuinely out of VRAM. The quantization guide covers the tradeoffs in detail.

Frequently asked questions

Can a Apple M3 Max 128GB run Command R 35B?

Yes — Command R 35B fits on a Apple M3 Max 128GB at FP16 (full) (84 GB of the 92 GB usable), with roughly 5 tokens/sec. Higher-precision quants need more VRAM; the table above shows each option.

How much VRAM does Command R 35B need?

Command R 35B is a 35B-parameter model. At FP16 that's about 84 GB; at Q4_K_M (the popular default) about 24 GB, including ~20% for the KV cache and runtime. Quantization is the main lever — see the per-quant table above.

Other combinations

Command R 35B on other GPUs: RTX 3060 12GB, RTX 4070 Ti, RTX 4080, RTX 3090, RTX 4090, NVIDIA L4

Other models on the Apple M3 Max 128GB: Mixtral 8x7B, Llama 3.3 70B, Qwen2.5 72B, Mixtral 8x22B, Llama 3.1 405B

Related