Jina OCR vs Typhoon OCR
Pick Typhoon OCR if you process Thai or Thai-English documents such as government forms, financial reports and books, or if you need a commercial-friendly licence. Pick Jina OCR for non-commercial or research PDF-to-Markdown work, where its 3B MoE model with about 570M active parameters scores 91.14 on OmniDocBench v1.6. The main split is licence and language focus: Jina is CC BY-NC 4.0 and needs a separate deal for commercial use, while Typhoon is Apache-2.0 under the OpenTyphoon terms and is built for Thai and English.
Jina OCRjina/jina-ocr | Typhoon OCRtyphoon/typhoon-ocr | |
|---|---|---|
| Type | Open source | Open source |
| Price | 1 credit / page | 2 credits / page |
| Per 1,000 pages | $1.00 | $2.00 |
| Typical speed | 1.5 s / page | 1.8 s / page |
| Best for | Clean Markdown at low cost | Thai documents |
| Inputs | PDF, images | PDF, images |
| Languages | About 100 languages | Thai, English |
| Max pages per request | 200 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | No |
| Math | ✓ Yes | No |
| Bounding boxes | No | No |
| Confidence | No | No |
| Forms | No | No |
| 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 Jina OCR if
- You are doing non-commercial or research work and want PDF to Markdown, HTML or LaTeX output.
- You want the higher OmniDocBench v1.6 score as published by the authors: 91.14 against 90.97 for Typhoon OCR V2 2B.
- You need throughput on a GPU: only about 570M parameters are active, and it runs at 2.57 pages per second on an A100.
Choose Typhoon OCR if
- You work with Thai government forms, financial reports, Thai books or mixed Thai-English documents.
- You need tables as HTML with merged cells, equations as LaTeX and checkboxes marked in the output.
- You want a smaller model: 2B parameters, with a 0.8B V2 variant, and an olmOCR-bench score of 84.7 against 83.4 for Jina OCR.
Key differences
| Jina OCR | Typhoon OCR | |
|---|---|---|
| Licence | CC-BY-NC-4.0. Commercial use needs a licence from Jina/Elastic. | Apache-2.0, subject to OpenTyphoon Terms and Conditions. |
| How you run it | Self-host the weights or call Jina's hosted API. The hosted API returns HTTP 503 on cold start; retry after 30 to 60 seconds. | Self-host the weights or use the Typhoon API as typhoon-ocr-v2. API pricing is not listed on the model card. |
| Model size | 3B MoE, about 570M active parameters. | 2B (V2 and V1.5), plus a 0.8B V2 variant. |
| OmniDocBench v1.6 | 91.14 (authors) | 90.97 for V2 2B (authors) |
| olmOCR-Bench | 83.4 (authors) | 84.7 for V2 2B (authors) |
| Languages | Not published | Thai and English. Only these two were evaluated. |
| Known failure | Reported scores use a stricter prompt than the recommended default. | V2 can loop until the token limit, mostly on dense multi-column pages. It can also output text that is not in the image. |
Sources
What each one is
Jina OCR
Compact page-to-Markdown model from Jina AI, built on the DeepSeek-OCR encoder.
Jina OCR detailsTyphoon OCR
Open vision model tuned for Thai and English documents, with Markdown output.
Typhoon 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": "jina/jina-ocr","fallbacks": ["typhoon/typhoon-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Jina OCR alternativesTyphoon OCR alternativesBest OCR models in 2026