Yes, just — 0.9 GB spare

Qwen3.5 122B-A10B at Q4_K_M fits your M2 Max · 96 GB entirely in unified memory at 8K context, at an estimated 19 tokens per second. Past 47K the KV cache pushes it over — quantise the cache to q8_0, or step down a quantisation, to go longer.

Fully in unified memory 8K context Q4_K_M · 70.3 GB Apache 2.0 Released 24 Feb 2026 Vision

The 96–128 GB unified-memory model: 122B of knowledge at 10B-active speed.

What hardware do I need for Qwen3.5 122B-A10B? →

The VRAM budget

weights 70.3 GB

Weights 70.3 GB KV cache @ 8K 0.19 GB Runtime overhead 0.6 GB Free 0.9 GB of 72.0 GB

Quantisation ladder

Quant	Weights	Total @ 8K	Max context	Tok/s	Quality	Fit
Q8_0	123.7 GB	124.5 GB	—	~11	−0.1% ppl	52.5 GB over
Q6_K	95.5 GB	96.2 GB	—	~14	−0.4% ppl	24.2 GB over
Q5_K_M	82.5 GB	83.3 GB	—	~16	−0.8% ppl	11.3 GB over
Q4_K_M	70.3 GB	71.1 GB	47K	19	−1.9% ppl	Recommended
Q3_K_M	56.9 GB	57.7 GB	256K	23	−5.4% ppl	Long context
Q2_K	48.7 GB	49.5 GB	256K	27	−15% ppl	Long context

Quality is the published perplexity delta against f16 weights. Max context assumes an f16 KV cache; q8_0 roughly doubles it. Only 12 of its 48 blocks keep a per-token KV cache; the rest are linear-attention, Mamba or convolution blocks with a fixed-size state.

How to run it

Ollama llama.cpp LM Studio MLX

terminal

$ pip install mlx-lm
$ mlx_lm.generate --model mlx-community/Qwen3.5-122B-A10B-4bit \
    --max-tokens 512 --prompt "Hello"

Apple's own array framework. The fastest path on Apple Silicon. More on MLX.

01Download is 70.3 GB. Keep it on an SSD — a first load off a spinning disk takes minutes.

02macOS caps what the GPU may wire down at about 72.0 GB of 96 GB. Raising it with iogpu.wired_limit_mb is possible, and risky.

03Only 0.9 GB is spare, so a long system prompt can still push it over. Its real ceiling here is 47K context.

See all models for this rig Compare with another model