RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can DeepSeek OCR run on the RTX 4090?

Yes. At 8K context the best fit is F16, using about 6.6 GB of the 22.8 GB usable on the RTX 4090. Estimated speed ceiling: ~502 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctxMax tok/s (est., 8K)
2-bit (Q2_K)2.0 GB Fits2.3 GB Fits2356
3-bit (Q3_K_M)2.4 GB Fits2.6 GB Fits1808
4-bit (Q4_K_M) •2.7 GB Fits2.9 GB Fits1511
5-bit (Q5_K_M)3.1 GB Fits3.3 GB Fits1240
6-bit (Q6_K)3.4 GB Fits3.7 GB Fits1077
8-bit (Q8_0) •3.8 GB Fits4.0 GB Fits943
16-bit (F16/BF16) •6.4 GB Fits6.6 GB Fits502

• real GGUF file size from sabafallah/DeepSeek-OCR-GGUF. Speed is a bandwidth ceiling (1008 GB/s), not a benchmark. Methodology.

More

DeepSeek OCR on every GPU and Mac · Every model that runs on the RTX 4090 · Other models that fit 24 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (deepseek-ai/DeepSeek-OCR), NVIDIA specs. Data as of 2026-11-17.