Model · Baidu · Vision
PaddleOCR 3.7
Baidu's open-source OCR toolkit. Version 3.7 ships PP-OCRv6, a classic text-detection and recognition pipeline in tiny, small and medium sizes, alongside PaddleOCR-VL-1.6, a 0.9B vision-language model for full document parsing into Markdown or JSON.
- Modality
- Vision
- License
- Apache 2.0 (Open)
- Parameter size
- PP-OCRv6: tiny 1.5M / small 7.7M / medium 34.5M; PaddleOCR-VL-1.6: 0.9B
- Released
- June 11, 2026
- Last verified
- October 28, 2026
- Runs locally
- Yes
What you need to run it
- Mac
- Yes. Any Apple Silicon Mac (M1 or newer) with at least 8 GB (16 GB recommended) of memory. Newer chips are faster, but memory is what decides whether it runs.
- Windows / Linux PC
- NVIDIA works best. A graphics card with at least 4 GB (6 GB recommended) of video memory.
- No graphics card
- Works. Runs on a normal computer's processor, just slower.
The classic PP-OCRv6 reader runs well on a plain CPU or an Apple Silicon Mac; the larger PaddleOCR-VL model is heavier.
Numbers are for: PP-OCRv6 classic text-reading pipeline. Our estimate from the model’s size — the maker publishes no figure.
Strengths
- Apache 2.0 — free for commercial use with no extra terms
- Tiny models run on an ordinary CPU or a phone; 5.2× faster end-to-end on CPU than PP-OCRv5 with OpenVINO
- One PP-OCRv6 model reads 50 languages (Chinese, English, Japanese and 46 Latin-script languages); the toolkit covers 100+
- PP-StructureV3 turns PDFs into Markdown or JSON with table-cell and text coordinates
- Models export to ONNX and run on ONNX Runtime, OpenVINO or TensorRT
Weaknesses
- Built on Baidu's PaddlePaddle framework rather than PyTorch, which is less familiar to most teams
- Two separate tracks (PP-OCRv6 pipeline vs PaddleOCR-VL model) — you have to pick the right one for the job
- Accuracy claims (beating large VLMs) are Baidu's own benchmarks
Try it
| Where | Type | Notes |
|---|---|---|
| GitHub | local | pip install; CPU, GPU and other accelerators |
| Hugging Face (PP-OCRv6) | weights | Apache 2.0 |
| Hugging Face (PaddleOCR-VL-1.6) | weights | Document-parsing VLM |
| PaddleOCR website | hosted | Online demo and docs |
Official sources
- GitHub github
- Documentation docs
Change log
- — Added hardware requirements (Mac support, memory, NVIDIA). Source: estimate from size and community builds (github PaddlePaddle/PaddleOCR README (mac/linux/win, CPU, M4 speed-up note)).
- — Initial entry. PaddleOCR 3.7.0 released 2026-06-11 (GitHub release) with PP-OCRv6 and PaddleOCR-VL-1.6, under Apache 2.0. Added 2026-10-28 from the popularity sweep: 91.8k GitHub stars.
Esc