Can I Run DeepSeek-OCR on Beelink SER9 (Ryzen AI 9, 32 GB)?

Written by Jakub Rusinowski · Last updated September 11, 2026

Beelink SER9 (Ryzen AI 9, 32 GB) — what it gives a model

Usable memory for models	32 GB
Memory bandwidth	256 GB/s
Form factor	Mini PC
Operating system	Windows or Linux
Memory upgradeable	Yes
Price	$859 (lib/data/ai-stations.ts (street price), checked 2026-07-06)

DeepSeek-OCR on Beelink SER9 (Ryzen AI 9, 32 GB): memory by quantization

Quant	Memory needed	Fits 32 GB?	Max context	Est. speed	Download
F16	7.6 GB	✓ Yes	8K	~96.5 tok/s	6 GB
Q8_0	4.8 GB	✓ Yes	8K	~131.8 tok/s	3.2 GB
Q6_K	4.1 GB	✓ Yes	8K	~145.7 tok/s	2.5 GB
Q5_K_M	3.8 GB	✓ Yes	8K	~153 tok/s	2.1 GB
Q4_K_M	3.5 GB	✓ Yes	8K	~160.7 tok/s	1.8 GB
Q3_K_M	2.9 GB	✓ Yes	8K	~175.6 tok/s	1.3 GB
Q2_K	2.6 GB	✓ Yes	8K	~185 tok/s	1 GB

What to watch out for

• This model does not publish its full attention configuration, so the KV-cache share of these figures is inferred from its parameter count rather than computed exactly.

Beelink SER9 32 GB limitations

• Shares system memory with the iGPU, so the usable model budget is well below the nominal 32 GB.
• Memory bandwidth, not capacity, is the limit here — expect single-digit tokens/sec on large models.

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.
• 32 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 inferred from the parameter count because this model does not publish its full attention configuration.

FAQ

Can I run DeepSeek-OCR on Beelink SER9 (Ryzen AI 9, 32 GB)?

Yes, comfortably — DeepSeek-OCR 3B at Q8_0 needs about 4.8 GB of the 32 GB usable on Beelink SER9 (Ryzen AI 9, 32 GB), leaving ~27.2 GB spare and running at ~131.8 tok/s (estimated), with room for about 8,192 tokens of context.

Which quantization of DeepSeek-OCR should I use on Beelink SER9 (Ryzen AI 9, 32 GB)?

Q8_0 — it needs about 4.8 GB of the 32 GB available, downloads as roughly 3.2 GB, and runs at an estimated 131.8 tokens/sec with up to 8K of context.

What limits DeepSeek-OCR on Beelink SER9 (Ryzen AI 9, 32 GB)?

Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.

Which runtime should I use?

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

Other Computers

• DeepSeek-OCR on Mac mini M4 16 GB
• DeepSeek-OCR on Mac mini M4 Pro 64 GB
• DeepSeek-OCR on Framework Desktop 128 GB

Other Models on Beelink SER9 (Ryzen AI 9, 32 GB)

• DeepSeek R1 on Beelink SER9 (Ryzen AI 9, 32 GB)
• Devstral on Beelink SER9 (Ryzen AI 9, 32 GB)
• EXAONE 3.5 on Beelink SER9 (Ryzen AI 9, 32 GB)
• Falcon 3 on Beelink SER9 (Ryzen AI 9, 32 GB)
• Gemma 2 Family on Beelink SER9 (Ryzen AI 9, 32 GB)

DeepSeek-OCR on GPUs

• DeepSeek-OCR on NVIDIA GeForce RTX 5060 Ti 8GB
• DeepSeek-OCR on NVIDIA GeForce RTX 5060
• DeepSeek-OCR on NVIDIA GeForce RTX 4060

What This Model Is Good At

• Best local LLMs for rag
• Best local LLMs for document analysis
• Best local LLMs for summarization

Model & Tools

• DeepSeek-OCR model page
• VRAM calculator
• Check your own hardware

← Can I Run It? | DeepSeek-OCR model page | Check your hardware