RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can LightOnOCR 3 0.8B run on the RTX 4070?

Yes. At 8K context the best fit is F16, using about 2.4 GB of the 11.4 GB usable on the RTX 4070. Estimated speed ceiling: ~332 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K)1.2 GB Fits1.3 GB Fits1.5 GB Fits2.7 GB Fits1576
3-bit (Q3_K_M) •1.4 GB Fits1.4 GB Fits1.7 GB Fits2.8 GB Fits1081
4-bit (Q4_K_M) •1.4 GB Fits1.5 GB Fits1.7 GB Fits2.9 GB Fits952
5-bit (Q5_K_M) •1.5 GB Fits1.5 GB Fits1.8 GB Fits2.9 GB Fits872
6-bit (Q6_K) •1.5 GB Fits1.6 GB Fits1.8 GB Fits3.0 GB Fits800
8-bit (Q8_0) •1.7 GB Fits1.7 GB Fits2.0 GB Fits3.1 GB Fits621
16-bit (F16/BF16) •2.3 GB Fits2.4 GB Fits2.7 GB Fits3.8 GB Fits332

• real GGUF file size from prithivMLmods/LightOnOCR-3-0.8B-GGUF. Speed is a bandwidth ceiling (504 GB/s), not a benchmark. Methodology.

More

LightOnOCR 3 0.8B on every GPU and Mac · Every model that runs on the RTX 4070 · Other models that fit 12 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (lightonai/LightOnOCR-3-0.8B), NVIDIA specs. Data as of 2026-11-17.