Qianfan OCR vs Nanonets-OCR
Pick Qianfan OCR for self-hosted multilingual parsing, tables, formulas and key information extraction, where it has an Apache-2.0 licence and a published OmniDocBench v1.5 score of 93.12. Pick Nanonets-OCR for printed business documents that need signatures, watermarks and checkboxes tagged, or flowcharts turned into Mermaid. Nanonets-OCR scores 69.5 on olmOCR-Bench against 79.8 for Qianfan OCR, so it is the weaker general parser.
Qianfan OCRbaidu/qianfan-ocr | Nanonets-OCRnanonets/nanonets-ocr | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 2 credits / page | 2 credits / page |
| Per 1,000 pages | $2.00 | $2.00 |
| Typical speed | 2.0 s / page | 1.8 s / page |
| Best for | Chinese documents | Markdown with semantic tags |
| Inputs | PDF, images | PDF, images |
| Languages | Chinese, English and more | Multilingual |
| Max pages per request | 200 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | No |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | No | No |
| Confidence | No | No |
| Forms | No | ✓ Yes |
| Multilingual | ✓ Yes | ✓ Yes |
| On OCRRouter | Available | Available |
Prices are what you pay on OCRRouter. Speed is the provider's typical figure, not a benchmark we ran.
Which to choose
Choose Qianfan OCR if
- It scores 79.8 on olmOCR-Bench, against 69.5 for Nanonets-OCR2-3B.
- It covers 192 languages, handwriting and scene text, while Nanonets-OCR was not trained on handwriting and can hallucinate.
- It is Apache-2.0 licensed, so commercial use is allowed, and it can also be used hosted on the Baidu Qianfan platform.
Choose Nanonets-OCR if
- It tags signatures, watermarks and image descriptions, and converts checkboxes to unicode symbols.
- It extracts flowcharts and org charts as Mermaid code, which Qianfan OCR does not list as an output.
- It is a 3B model fine-tuned from Qwen2.5-VL-3B, and it distinguishes inline and display LaTeX equations.
Key differences
| Qianfan OCR | Nanonets-OCR | |
|---|---|---|
| Licence | Apache-2.0 | Not stated on the Nanonets-OCR2-3B model card (no license tag) |
| How you run it | Self-host the open weights, or use the Baidu Qianfan platform | Self-host the weights from Hugging Face only |
| Output formats | Markdown, JSON, HTML, LaTeX, bounding boxes | Markdown, HTML, LaTeX, Mermaid |
| olmOCR-Bench | 79.8 | 69.5 (OCR2-3B) |
| Handwriting | Listed as supported, along with scene text | Not trained on handwriting. One handwriting test saw generation loops with CER up to 2100% |
| Prompts | Thinking mode can emit up to 16,384 tokens, adding latency | Trained on a fixed prompt. The team says hallucination with other prompts is expected |
| Speed and hardware | 1.024 pages per second on one A100 with W8A8. Not suited to CPU-only deployment | Not published |
Sources
- Qianfan-OCR model card
- Qianfan-OCR paper
- Nanonets-OCR2-3B model card
- Nanonets-OCR-s announcement
- HF discussion fixed prompt
- Japanese handwriting OCR comparison
What each one is
Qianfan OCR
Baidu's OCR-tuned vision model. Strong on Chinese documents and tables.
Qianfan OCR detailsNanonets-OCR
Open image-to-Markdown model that tags tables, equations, signatures, watermarks and checkboxes.
Nanonets-OCR detailsTry both on your own file
Run the file once with each model and compare the output.
Use one, fall back to the other
{"model": "baidu/qianfan-ocr","fallbacks": ["nanonets/nanonets-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Qianfan OCR alternativesNanonets-OCR alternativesBest OCR models in 2026