RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can olmOCR 2 7B 1025 run on the L40S?

Yes. At 8K context the best fit is F16, using about 15.3 GB of the 45.6 GB usable on the L40S. Estimated speed ceiling: ~57 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •3.7 GB Fits3.9 GB Fits5.2 GB Fits286
3-bit (Q3_K_M) •4.5 GB Fits4.7 GB Fits6.0 GB Fits227
4-bit (Q4_K_M) •5.3 GB Fits5.5 GB Fits6.8 GB Fits184
5-bit (Q5_K_M) •6.0 GB Fits6.2 GB Fits7.5 GB Fits159
6-bit (Q6_K) •6.7 GB Fits7.0 GB Fits8.3 GB Fits138
8-bit (Q8_0) •8.5 GB Fits8.7 GB Fits10.0 GB Fits107
16-bit (F16/BF16) •15.1 GB Fits15.3 GB Fits16.6 GB Fits57

• real GGUF file size from bartowski/allenai_olmOCR-2-7B-1025-GGUF. Speed is a bandwidth ceiling (864 GB/s), not a benchmark. Methodology.

More

olmOCR 2 7B 1025 on every GPU and Mac · Every model that runs on the L40S · Other models that fit 48 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (allenai/olmOCR-2-7B-1025), NVIDIA specs. Data as of 2026-11-17.