OCR and documents · Baidu PaddlePaddle
PaddleOCR-VL
0.9B document parser covering 109 languages.
Datasheet
Maker | Baidu PaddlePaddle |
|---|---|
Parameters | 959 MVerifiedsafetensors.total from the HF API Source |
Licence | Apache-2.0Verified |
Access | Public download |
Task (HF pipeline) | image-text-to-text |
First published on HF | 2025-10-16 |
Last updated on HF | 2026-08-08 |
Weights in memory | 1.9 GB at 16-bit · 0.6 GB at ~4.85 bitsEstimateParameters × bits ÷ 8. Excludes context cache and runtime overhead; not every model has a 4-bit build. |
Benchmark
80%
olmOCR-Bench, overall · #2 of 4 scored in this category
PaddleOCR-VL* row (±1.0); asterisk as marked in the source table. Source
Used in
The description summarises the model card at https://huggingface.co/PaddlePaddle/PaddleOCR-VL. Parameters and licence are read from the Hugging Face API; the weights-in-memory line is our arithmetic, not a measurement.