Qwen · Mixture-of-experts · Apache 2.0
Can I run Qwen3 235B-A22B (MoE) locally?
Frontier-class open weights. ~140 GB at Q4 — a 192 GB Mac Studio or a multi-GPU server. Fast for its size.
- Parameters
- 235B (22B active)
- VRAM at Q4
- 140 GB
- Max context
- 131K
- Released
- 2025-04
Memory needed by quantisation
Total includes weights, an 8K-token KV cache and runtime overhead. Lower quants trade quality for size.
| Quant | Bits/weight | Weights | Total needed | Quality |
|---|---|---|---|---|
| FP16 | 16 | 438 GB | 457 GB | Full precision. Reference quality, 2× the size of 8-bit. |
| Q8_0 | 8.5 | 233 GB | 244 GB | Effectively lossless. The safe choice when it fits. |
| Q6_K | 6.56 | 179 GB | 189 GB | Near-lossless; quality loss is hard to measure. |
| Q5_K_M | 5.67 | 155 GB | 164 GB | Very good. A common sweet spot above Q4. |
| Q4_K_M | 4.83 | 132 GB | 140 GB | The default. Best size/quality trade-off for local use. |
| Q3_K_M | 3.91 | 107 GB | 113 GB | Noticeable degradation; useful to squeeze a size up. |
| Q2_K | 3.35 | 92 GB | 98 GB | Aggressive. Quality drops a lot — last resort to fit. |
Which hardware runs Qwen3 235B-A22B (MoE)?
Best quantisation that fits each device at 8K context, with a rough speed estimate. Try your exact setup →
| Hardware | Memory | Fit | Best quant | Speed |
|---|---|---|---|---|
| Mac Studio · M3 Ultra, 256 GB | 256 GB | Runs well | Q5 | ~25 tok/s |
| Mac Studio · M3 Ultra, 512 GB | 512 GB | Runs well | Q8 | ~16 tok/s |
| CPU only · 128 GB RAM | 128 GB | Runs (tight) | Q3 | ~3.9 tok/s |
| RTX 3060 12 GB | 12 GB | Won't fit | — | — |
| RTX 4060 Ti 8 GB | 8 GB | Won't fit | — | — |
| RTX 4060 Ti 16 GB | 16 GB | Won't fit | — | — |
| RTX 3080 10 GB | 10 GB | Won't fit | — | — |
| RTX 5070 | 12 GB | Won't fit | — | — |
| RTX 4070 Super | 12 GB | Won't fit | — | — |
| Radeon RX 7900 XT | 20 GB | Won't fit | — | — |
| RTX 5070 Ti | 16 GB | Won't fit | — | — |
| RTX 4070 Ti Super | 16 GB | Won't fit | — | — |
| RTX 3090 | 24 GB | Won't fit | — | — |
| Radeon RX 7900 XTX | 24 GB | Won't fit | — | — |
| RTX 5080 | 16 GB | Won't fit | — | — |
| RTX 4080 Super | 16 GB | Won't fit | — | — |
| RTX 4090 | 24 GB | Won't fit | — | — |
| RTX 5090 | 32 GB | Won't fit | — | — |
| RTX A6000 | 48 GB | Won't fit | — | — |
| RTX 6000 Ada | 48 GB | Won't fit | — | — |
| A100 80 GB | 80 GB | Won't fit | — | — |
| H100 80 GB | 80 GB | Won't fit | — | — |
| Mac · M1/M2/M3 (base), 8 GB | 8 GB | Won't fit | — | — |
| Mac · M1/M2/M3 (base), 16 GB | 16 GB | Won't fit | — | — |
| Mac · M4 (base), 24 GB | 24 GB | Won't fit | — | — |
| Mac · M4 Pro, 48 GB | 48 GB | Won't fit | — | — |
| Mac · M1/M2/M3 Max, 32 GB | 32 GB | Won't fit | — | — |
| Mac · M4 Max, 64 GB | 64 GB | Won't fit | — | — |
| Mac · M1/M2/M3 Max, 64 GB | 64 GB | Won't fit | — | — |
| Mac · M3/M4 Max, 128 GB | 128 GB | Won't fit | — | — |
| Mac Studio · M1/M2 Ultra, 128 GB | 128 GB | Won't fit | — | — |
| CPU only · 8 GB RAM | 8 GB | Won't fit | — | — |
| CPU only · 16 GB RAM | 16 GB | Won't fit | — | — |
| CPU only · 32 GB RAM | 32 GB | Won't fit | — | — |
| CPU only · 64 GB RAM | 64 GB | Won't fit | — | — |
What Qwen3 235B-A22B (MoE) is good for
Related models
FAQ
How much VRAM does Qwen3 235B-A22B (MoE) need?
At Q4_K_M, Qwen3 235B-A22B (MoE) needs about 140 GB including a 8K-token context and overhead (132 GB for the weights alone). Higher quantisation needs more; see the table for every level.
What is the cheapest way to run Qwen3 235B-A22B (MoE)?
The smallest device that runs it well in our list is the Mac Studio · M3 Ultra, 256 GB (256 GB). Anything with at least that much memory should handle it at a usable quantisation.
Is Qwen3 235B-A22B (MoE) good for general chat & assistance?
Frontier-class open weights. ~140 GB at Q4 — a 192 GB Mac Studio or a multi-GPU server. Fast for its size.
Estimates — see how we compute these. Memory figures assume an 8K context; long-context use needs more.