Can I Run Qwen 2.5 VL on Framework Desktop (Ryzen AI Max+ 395, 128 GB)?

Superseded model. Qwen 2.5 VL has been superseded by Qwen 3. This page is kept for reference; the newer family is a better starting point. View Qwen 3 →

Written by Jakub Rusinowski · Last updated January 25, 2025

Yes — comfortably

Yes, comfortably — Qwen 2.5 VL 72B Instruct at Q3_K_M needs about 34.8 GB of the 128 GB usable on Framework Desktop (Ryzen AI Max+ 395, 128 GB), leaving ~93.2 GB spare and running at ~5.8 tok/s (estimated), with room for about 65,536 tokens of context.

Confidence: high · Recommended quantization: Q3_K_M · Estimated speed: ~5.8 tok/s

Framework Desktop (Ryzen AI Max+ 395, 128 GB) — what it gives a model

Usable memory for models	128 GB
Memory bandwidth	256 GB/s
Form factor	Mini PC
Operating system	Windows or Linux
Memory upgradeable	No — soldered

Qwen 2.5 VL on Framework Desktop (Ryzen AI Max+ 395, 128 GB): memory by quantization

Quant	Memory needed	Fits 128 GB?	Max context	Est. speed	Download
F16	150.3 GB	✗ No	—	—	146.8 GB
Q8_0	81.5 GB	✓ Yes	64K	~2.4 tok/s	78 GB
Q6_K	63.7 GB	✓ Yes	64K	~3.1 tok/s	60.2 GB
Q5_K_M	55.5 GB	✓ Yes	64K	~3.6 tok/s	52 GB
Q4_K_M	47.8 GB	✓ Yes	64K	~4.2 tok/s	44.3 GB
Q3_K_M	34.8 GB	✓ Yes	64K	~5.8 tok/s	31.3 GB
Q2_K	27.6 GB	✓ Yes	64K	~7.4 tok/s	24.1 GB

Which Qwen 2.5 VL sizes fit

Variant	Needs at Q4_K_M	Fits?	Est. speed
Qwen 2.5 VL 72B Instruct	47.8 GB	✓ Fits	~4.2 tok/s
Qwen 2.5 VL 7B Instruct	6.3 GB	✓ Fits	~33.9 tok/s

What to watch out for

Q3_K_M is the only quantization that fits, and it is a heavily degraded one — expect noticeably worse output than the same model at Q4_K_M. A smaller model at Q4 is usually the better trade.
Memory on this machine is not upgradeable, so the configuration you buy is the ceiling for every model you will ever run on it.

Framework Desktop 128 GB limitations

Memory is soldered LPDDR5X — unusually for Framework, this is the one component that cannot be upgraded.
ROCm/Vulkan support for Strix Halo is younger than CUDA; check your runtime supports it before buying.

Recommended setup

Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving

How these numbers are calculated

Memory need = quantized weights + KV cache at 8,192 tokens (f16) + 0.8 GB runtime overhead.
128 GB unified memory at 256 GB/s, shared between CPU and GPU.
Throughput is a memory-bandwidth roofline estimate, not a measurement. It is labelled "estimated" everywhere it appears.
KV cache is computed from this model's published attention configuration.

FAQ