Written by Jakub Rusinowski · Last updated April 28, 2025
Yes — comfortably
Yes, comfortably — Qwen 3 32B at Q8_0 needs about 37.8 GB of the 128 GB usable on Framework Desktop (Ryzen AI Max+ 395, 128 GB), leaving ~90.2 GB spare and running at ~5.3 tok/s (estimated), with room for about 65,536 tokens of context.
Confidence: high · Recommended quantization: Q8_0 · Estimated speed: ~5.3 tok/s
| Usable memory for models | 128 GB |
| Memory bandwidth | 256 GB/s |
| Form factor | Mini PC |
| Operating system | Windows or Linux |
| Memory upgradeable | No — soldered |
| Quant | Memory needed | Fits 128 GB? | Max context | Est. speed | Download |
|---|---|---|---|---|---|
| F16 | 68.5 GB | ✓ Yes | 64K | ~2.9 tok/s | 65.6 GB |
| Q8_0 | 37.8 GB | ✓ Yes | 64K | ~5.3 tok/s | 34.8 GB |
| Q6_K | 29.8 GB | ✓ Yes | 64K | ~6.8 tok/s | 26.9 GB |
| Q5_K_M | 26.2 GB | ✓ Yes | 64K | ~7.8 tok/s | 23.2 GB |
| Q4_K_M | 22.8 GB | ✓ Yes | 64K | ~9 tok/s | 19.8 GB |
| Q3_K_M | 16.9 GB | ✓ Yes | 64K | ~12.4 tok/s | 14 GB |
| Q2_K | 13.7 GB | ✓ Yes | 64K | ~15.6 tok/s | 10.8 GB |
| Variant | Needs at Q4_K_M | Fits? | Est. speed |
|---|---|---|---|
| Qwen 3 235B-A22B (MoE) | 144.3 GB | ✗ Too large | — |
| Qwen 3 32B | 22.8 GB | ✓ Fits | ~9 tok/s |
| Qwen 3 30B-A3B (MoE) | 20 GB | ✓ Fits | ~67.9 tok/s |
| Qwen 3 14B | 11.1 GB | ✓ Fits | ~19.1 tok/s |
| Qwen 3 8B | 7 GB | ✓ Fits | ~32.1 tok/s |
Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving
Yes, comfortably — Qwen 3 32B at Q8_0 needs about 37.8 GB of the 128 GB usable on Framework Desktop (Ryzen AI Max+ 395, 128 GB), leaving ~90.2 GB spare and running at ~5.3 tok/s (estimated), with room for about 65,536 tokens of context.
Q8_0 — it needs about 37.8 GB of the 128 GB available, downloads as roughly 34.8 GB, and runs at an estimated 5.3 tokens/sec with up to 64K of context.
Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.
Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving