Can you run Qwen3 235B A22B locally?
Qwen3 family · · 235.1B parameters (22B activated per token) · Hugging Face model card
Qwen3 235B A22B is a mixture-of-experts model with 22.0B active parameters per token and 235.1B total open-weight model with a 40,960-token context window, usable locally with roughly 192 GB of memory at Q4-class quantization (calculated; measured speeds vary by hardware).
Minimum: 192 GB+ memory (Q4_K_M, calculated) · Recommended: 192 GB+ memory for full context (Q4_K_M, calculated)
How much VRAM does Qwen3 235B A22B need at each quantization?
Qwen3 235B A22B needs 156.0 GB of VRAM at Q4_K_M. The table below lists weights-only size and total VRAM including overhead for each common quantization level with 4k and 32k token contexts.
| Quantization | Bits / weight | Weights only | Total + KV @4k ctx | Total + KV @32k ctx |
|---|---|---|---|---|
| Q4_K_M | 4.8 | 141.1 GB | 156.0 GB | 161.5 GB |
| Q5_K_M | 5.7 | 167.5 GB | 185.0 GB | 190.6 GB |
| Q6_K | 6.6 | 194.0 GB | 214.1 GB | 219.7 GB |
| Q8_0 | 8.5 | 249.8 GB | 275.6 GB | 281.1 GB |
| FP16 | 16 | 470.2 GB | 518.0 GB | 523.5 GB |
Method: weights = parameters × bits-per-weight (Q4_K_M ≈ 4.8, Q5_K_M ≈ 5.7, Q6_K ≈ 6.6, Q8_0 ≈ 8.5, FP16 = 16), plus 10% loading overhead, plus KV cache from the verified architecture config (94 layers, 4 KV heads, 128 head dim). Source: HF-verified 2026-09-12 via api (params=235.09B ctx=40960 license=apache-2.0); VRAM figures are documented calculations, not measurements
Which GPUs can run Qwen3 235B A22B locally?
At Q4_K_M with a 4k context, Qwen3 235B A22B needs 156.0 GB of VRAM. The lists below are computed live from our GPU database and grouped by how much headroom the card has. Disclosure: CompareAIHardware.com participates in the Amazon Associates program and earns from qualifying purchases through links on this page. Affiliate relationships do not influence our recommendations.
Runs Qwen3 235B A22B comfortably (25%+ VRAM headroom)
| GPU | VRAM | Bandwidth | Type | |
|---|---|---|---|---|
| AMD Instinct MI325X | 256 GB | 6000 GB/s | Data Center GPU | |
| AMD Instinct MI355X | 288 GB | 8000 GB/s | Data Center GPU | |
| NVIDIA B300 (Blackwell Ultra) | 288 GB | 8000 GB/s | Data Center GPU |
Minimum GPUs that fit Qwen3 235B A22B at Q4_K_M
These GPUs hold the model but leave little headroom — keep contexts short.
| GPU | VRAM | Bandwidth | Type | |
|---|---|---|---|---|
| AMD Instinct MI300X | 192 GB | 5300 GB/s | Data Center GPU | |
| NVIDIA B200 | 192 GB | 8000 GB/s | Data Center GPU |
Needs 2+ GPUs to run Qwen3 235B A22B
One of these cards is too small on its own, but a pair (tensor or pipeline parallel, ~90% efficiency) covers the 156.0 GB requirement. See our multi-GPU guide for setup.
| GPU | VRAM | Bandwidth | Type | |
|---|---|---|---|---|
| NVIDIA RTX PRO 6000 Blackwell | 96 GB ×2 | 1792 GB/s | Pro GPU | Check price |
| H200 SXM | 141 GB ×2 | 4800 GB/s | Data Center GPU |
Can a Mac run Qwen3 235B A22B?
Yes — these Apple Silicon machines fit Qwen3 235B A22B at Q4_K_M, since macOS lets the GPU use about 75% of unified memory. Generation speed is bound by memory bandwidth, so the GB/s column matters as much as capacity.
| Apple Silicon | Unified memory | Usable by GPU (~75%) | Bandwidth |
|---|---|---|---|
| M3 Ultra (Mac Studio) | 512 GB | 384 GB | 819 GB/s |
Which pre-built systems can run Qwen3 235B A22B?
These mini PCs and workstations from our database fit Qwen3 235B A22B at Q4_K_M. Usable-memory figures are conservative: Windows shares about half of system RAM with the GPU by default (Linux can expose more), macOS lets Apple Silicon GPUs use about 75% of unified memory, and Linux unified-memory systems such as GB10 expose roughly 90%. Generation speed is bound by memory bandwidth, so compare the GB/s column before buying.
| System | Type | Memory | Usable for AI | Bandwidth | GPUs |
|---|---|---|---|---|---|
| HP Z8 Fury G5 (4x RTX 6000 Ada) | Workstation | 192 GB | 192 GB | 960 GB/s | 4 |