Can I Run Qwen 3 235B-A22B (MoE) on Apple M2 Ultra?
作者： Jakub Rusinowski · 最后更新： 2026年7月12日
Not at Q4_K_M — that needs 144.3 GB at 8K context and the Apple M2 Ultra has 144 GB usable. Drop to Q3_K_M and it fits in 102.5 GB, with up to 64K of context.
Get a personalized upgrade path →
The numbers
| Model | Qwen 3 235B-A22B (MoE) |
| Parameters | 235B (22B active) |
| Quantization | Q4_K_M |
| VRAM needed (8K context) | 144.3 GB |
| Apple M2 Ultra usable memory | 144 GB |
| Verdict | Runs at a lower quant |
| Fits at | Q3_K_M (102.5 GB) |
| Max context | 64K |
Qwen 3 235B-A22B (MoE) on the Apple M2 Ultra: VRAM by quantization
| Quant | VRAM needed | Fits 192 GB? | Max context | Whole PC |
|---|---|---|---|---|
| F16 | 472.4 GB | ✗ No | — | — |
| Q8_0 | 252.1 GB | ✗ No | — | — |
| Q6_K | 195.1 GB | ✗ No | — | — |
| Q5_K_M | 168.9 GB | ✗ No | — | — |
| Q4_K_M | 144.3 GB | ✗ No | — | — |
| Q3_K_M | 102.5 GB | ✓ Yes | 64K | — |
| Q2_K | 79.6 GB | ✓ Yes | 64K | — |
Assumes an 8K-token context with an f16 KV cache. A longer window needs more; a quantized KV cache needs less. “Max context” is the largest window that still fits in 192 GB of unified memory. Figures are estimates from parameter count, quantization and memory bandwidth — the analyzer lets you tune KV-cache quant and context.
Context costs VRAM too. At Q4_K_M on the Apple M2 Ultra: 8K 144.3 GB ✗ · 32K 149 GB ✗ · 125K 167.3 GB ✗. Past 8K it no longer fits 192 GB — a q8_0 KV cache buys roughly half of that back, which the calculator will price for you.
What to do next
Quantize to Q3_K_M. That is the cheapest fix — it costs some quality but no money.
No compatible GPU? Qwen 3 235B-A22B (MoE) on 192 GB of system RAM, CPU only: runs comfortably.
Running Qwen 3 235B-A22B (MoE) on this card:
Qwen 3 235B-A22B (MoE) needs ~144 GB but this GPU has 192 GB. Rent a 2× A100 (160 GB)-class GPU by the hour instead of buying one:
Affiliate links — we may earn a commission if you sign up, at no extra cost to you.
Cloud rates verified 2026-07 — estimates, and marketplace prices vary. Buying price is GPU MSRP only, not a full PC.
Other Qwen 3 sizes on the Apple M2 Ultra
- Qwen 3 32B — runs (22.8 GB)
- Qwen 3 30B-A3B (MoE) — runs (20 GB)
- Qwen 3 14B — runs (11.1 GB)
- Qwen 3 8B — runs (7 GB)
Qwen 3 235B-A22B (MoE) on nearby hardware
- NVIDIA L40S (48 GB) — does not fit
All Qwen 3 sizes on the Apple M2 Ultra | Qwen 3 235B-A22B (MoE) model page | Apple M2 Ultra GPU page | VRAM calculator | Check your hardware