RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can GOT OCR2 0 run on the H100 SXM?

Yes. At 8K context the best fit is F16, using about 2.8 GB of the 76.0 GB usable on the H100 SXM. Estimated speed ceiling: ~2339 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctxMax tok/s (est., 8K)
2-bit (Q2_K)1.3 GB Fits1.7 GB Fits3.9 GB Fits12476
3-bit (Q3_K_M)1.4 GB Fits1.8 GB Fits4.0 GB Fits9572
4-bit (Q4_K_M)1.5 GB Fits1.8 GB Fits4.1 GB Fits7654
5-bit (Q5_K_M)1.5 GB Fits1.9 GB Fits4.2 GB Fits6566
6-bit (Q6_K)1.6 GB Fits2.0 GB Fits4.2 GB Fits5706
8-bit (Q8_0)1.8 GB Fits2.1 GB Fits4.4 GB Fits4403
16-bit (F16/BF16)2.4 GB Fits2.8 GB Fits5.0 GB Fits2339

• real GGUF file size. Speed is a bandwidth ceiling (3350 GB/s), not a benchmark. Methodology.

More

GOT OCR2 0 on every GPU and Mac · Every model that runs on the H100 SXM · Other models that fit 64 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (stepfun-ai/GOT-OCR2_0), NVIDIA specs. Data as of 2026-11-17.