olmOCR models
2 olmOCR models from Ai2 on Hugging Face, from 8.3B to 8.3B parameters, published by Ai2 in BF16, with a 125K-token context. The smallest official model, olmOCR-2-7B-1025, needs about 18.5 GB of VRAM at its published precision; the cheapest live fit is RTX A5000 at $0.176/hr.
Part of the OLMo series
Pick a size
One row per official olmOCR size: the VRAM it needs at each precision, the cheapest GPU that holds it today, and the KV cache for a 32K-token context.
| Model | Parameters | Native VRAM | FP8 VRAM | INT4 VRAM | Live GPU fit (native) | Est. $/hr | KV cache at 32K |
|---|---|---|---|---|---|---|---|
| olmOCR-2-7B-1025 | 8.3B | 18.5 GB | 9.3 GB | 4.6 GB | RTX A5000 | $0.176/hr | 1.8 GB |
VRAM is weights times a flat 1.2 overhead; the KV cache is a separate per-model figure at 16-bit, shown where the architecture is published. See the methodology.
Official models (1)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| olmOCR-2-7B-1025 | 8.3B | BF16 | 18.5 GB | RTX A5000 | $0.176/hr |
Quantized versions (1)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| olmOCR-2-7B-1025-FP8 | 8.3B | F8_E4M3 | 9.3 GB | RTX 4070 | $0.154/hr |
VRAM is for the precision the model is published in: the weights times a flat 1.2 overhead, with the KV cache not included. 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.