Typhoon OCR vs Tesseract.js
Pick Typhoon OCR for Thai and English documents such as government forms, financial reports and books, where you need tables as HTML, equations as LaTeX and Markdown output. Pick Tesseract.js for in-browser OCR of clean printed text in many languages when images must stay on the user's device. The main split is document understanding on a 2B model versus a plain-text engine that runs client side.
Typhoon OCRtyphoon/typhoon-ocr | Tesseract.jstesseract/tesseract-js | |
|---|---|---|
| Type | Open source | Open source |
| Price | 2 credits / page | 1 credit / page |
| Per 1,000 pages | $2.00 | $1.00 |
| Typical speed | 1.8 s / page | 2.0 s / page |
| Best for | Thai documents | OCR in the browser |
| Inputs | PDF, images | images |
| Languages | Thai, English | 100+ languages |
| Max pages per request | 500 | 500 |
| Tables | ✓ Yes | No |
| Handwriting | No | No |
| Math | No | No |
| Bounding boxes | No | ✓ Yes |
| Confidence | No | ✓ Yes |
| 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 Typhoon OCR if
- You work with Thai government forms, financial reports or Thai-English mixed documents, where V1.5 reached BLEU 0.870 and beat GPT-5 and Gemini 2.5 Pro on forms.
- You need structure in the output: tables as HTML with merged cells, equations as LaTeX, and checkboxes marked.
- You want a published benchmark record: V2 2B scores 84.7 on olmOCR-bench and 90.97 on OmniDocBench v1.6.
Choose Tesseract.js if
- You need OCR to run fully in the browser or Node.js, so images never leave the user's device.
- You need one of more than 100 languages, while Typhoon OCR was evaluated only on Thai and English.
- You only need plain text, hOCR or bounding boxes from clean screenshots or simple printed scans, installed from npm, ESM or a CDN script tag.
Key differences
| Typhoon OCR | Tesseract.js | |
|---|---|---|
| How you run it | Self-host the weights, or call the Typhoon API as typhoon-ocr-v2 | JavaScript and WebAssembly library in the browser or Node.js |
| Licence | Apache-2.0, subject to OpenTyphoon Terms and Conditions | Apache-2.0 |
| Price | Free self-hosted (you pay for compute). API pricing not listed on the model card | Free, runs on the user's device or your server |
| Output formats | Markdown, HTML, LaTeX | Plain text, hOCR, bounding boxes |
| Languages | Thai and English only evaluated | More than 100 via traineddata files |
| Benchmarks | ThaiOCRBench 59.9, olmOCR-bench 84.7, OmniDocBench v1.6 90.97 (V2 2B, author-reported) | Not published |
| Known failure | V2 can loop until the token limit on dense multi-column pages, and key-information extraction scores only 21.0 on ThaiOCRBench | No PDF input, lower accuracy than the native Tesseract CLI on the same image, and fails on screen resolution text unless upscaled |
Sources
- Typhoon OCR V2 2B model card
- Typhoon OCR technical report
- Typhoon OCR 1.5 release blog
- GitHub README
- Issue #251 accuracy vs CLI
What each one is
Typhoon OCR
Open vision model tuned for Thai and English documents, with Markdown output.
Typhoon OCR detailsTesseract.js
Tesseract compiled to JavaScript and WebAssembly, so OCR runs in the browser or in Node.js.
Tesseract.js 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": "typhoon/typhoon-ocr","fallbacks": ["tesseract/tesseract-js"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Typhoon OCR alternativesTesseract.js alternativesBest OCR models in 2026