lightonai models
3 lightonai models on Hugging Face, from 1.0B to 1.2B parameters. At the precision each one is published in, the smallest needs about 2.2 GB of VRAM (LightOnOCR-2-1B, cheapest live fit: RTX 5060 Ti) and the largest about 2.6 GB (LightOnOCR-1B-1025, cheapest live fit: RTX 5060 Ti). The cheapest way to run LightOnOCR-2-1B is $0.110/hr.
lightonai models
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| LightOnOCR-2-1B | 1.0B | BF16 | 2.2 GB | RTX 5060 Ti | $0.110/hr |
| LightOnOCR-2-1B-bbox-soup | 1.0B | BF16 | 2.2 GB | RTX 5060 Ti | $0.110/hr |
| LightOnOCR-1B-1025 | 1.2B | BF16 | 2.6 GB | RTX 5060 Ti | $0.110/hr |
VRAM is for the precision the model is published in, with the same overhead and an 8,192-token context assumed on every page; see the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
