Written by Jakub Rusinowski · Last updated September 11, 2026
Yes — comfortably
Yes, comfortably — DeepSeek-OCR 3B at Q8_0 needs about 4.8 GB of the 24 GB usable on RTX 4090 Desktop (24 GB VRAM, 64 GB RAM), leaving ~19.2 GB spare and running at ~277.6 tok/s (estimated), with room for about 8,192 tokens of context.
Confidence: medium · Recommended quantization: Q8_0 · Estimated speed: ~277.6 tok/s
or compare on Vast.ai from $0.35/hr (typical low · varies)
As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.
| Usable memory for models | 24 GB |
| Memory bandwidth | 1008 GB/s |
| Form factor | Desktop |
| Operating system | Windows or Linux |
| Memory upgradeable | Yes |
| Quant | Memory needed | Fits 24 GB? | Max context | Est. speed | Download |
|---|---|---|---|---|---|
| F16 | 7.6 GB | ✓ Yes | 8K | ~232.1 tok/s | 6 GB |
| Q8_0 | 4.8 GB | ✓ Yes | 8K | ~277.6 tok/s | 3.2 GB |
| Q6_K | 4.1 GB | ✓ Yes | 8K | ~292.5 tok/s | 2.5 GB |
| Q5_K_M | 3.8 GB | ✓ Yes | 8K | ~299.8 tok/s | 2.1 GB |
| Q4_K_M | 3.5 GB | ✓ Yes | 8K | ~307.1 tok/s | 1.8 GB |
| Q3_K_M | 2.9 GB | ✓ Yes | 8K | ~320.3 tok/s | 1.3 GB |
| Q2_K | 2.6 GB | ✓ Yes | 8K | ~328 tok/s | 1 GB |
Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving
Yes, comfortably — DeepSeek-OCR 3B at Q8_0 needs about 4.8 GB of the 24 GB usable on RTX 4090 Desktop (24 GB VRAM, 64 GB RAM), leaving ~19.2 GB spare and running at ~277.6 tok/s (estimated), with room for about 8,192 tokens of context.
Q8_0 — it needs about 4.8 GB of the 24 GB available, downloads as roughly 3.2 GB, and runs at an estimated 277.6 tokens/sec with up to 8K 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
← Can I Run It? | DeepSeek-OCR model page | Check your hardware