Chandra vs Typhoon OCR
Pick Chandra for multilingual scans, PDF to Markdown for RAG, and pages with tables, forms and math, when your revenue or funding is under $2M or you use the Datalab API. Pick Typhoon OCR for Thai and English documents such as government forms and financial reports, where it is free to self-host under Apache-2.0. The main split is language coverage: Chandra is benchmarked on 43 to 90 languages, Typhoon OCR is only evaluated on Thai and English.
Chandradatalab/chandra | Typhoon OCRtyphoon/typhoon-ocr | |
|---|---|---|
| Type | Open source | 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 | Complex layouts and handwriting | Thai documents |
| Inputs | PDF, images | PDF, images |
| Languages | 40+ languages | Thai, English |
| Max pages per request | 500 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | No |
| Math | ✓ Yes | No |
| Bounding boxes | ✓ Yes | No |
| Confidence | No | No |
| Forms | ✓ Yes | 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 Chandra if
- It scores 85.8 on olmOCR-bench (Chandra 2) and averages 77.8% across 43 languages.
- It outputs Markdown, HTML and JSON, plus layout chunks with bounding boxes and labels per block.
- You can use the hosted Datalab API at $1, $3 or $7.50 per 1,000 pages, or self-host on a single GPU such as an A100, L40S or RTX 4090/5090.
Choose Typhoon OCR if
- V1.5 beat GPT-5 and Gemini 2.5 Pro on Thai government forms, with a BLEU of 0.870.
- V2 2B scores 81.3 on MDPBench Thai pages versus 61.3 for its base model, and 84.7 on olmOCR-bench at only 2B parameters.
- The licence is Apache-2.0, subject to OpenTyphoon Terms and Conditions, with no revenue cap listed, and it produces HTML tables with merged cells, LaTeX equations and marked checkboxes.
Key differences
| Chandra | Typhoon OCR | |
|---|---|---|
| Licence | Code Apache-2.0; weights modified OpenRAIL-M. Free for research, personal use and companies under $2M funding/revenue; a commercial licence is needed above that. | Apache-2.0, subject to OpenTyphoon Terms and Conditions. |
| Languages | 77.8% average across 43 languages and 72.7% across 90 languages. Weak on Urdu (43.2%) and Sanskrit (51.1%). | Only Thai and English evaluated. |
| Price | Self-hosted is free within licence terms. Datalab API: $1, $3 or $7.50 per 1,000 pages, with $20/month free tier. | Self-hosted is free. Typhoon API pricing not published. |
| Model size and hardware | Single NVIDIA GPU via vLLM; benchmarked on one H100 80GB, about 1.44 pages/s. | 2B parameters (V2 and V1.5), with a 0.8B V2 variant. Hardware requirements not published. |
| Output formats | Markdown, HTML, JSON and bounding boxes. | Markdown, HTML and LaTeX. |
| olmOCR-bench | 85.8 (Chandra 2) | 84.7 (V2 2B) |
| Known failure | Token-loop hallucinations reported on 5-6 of about 60 Russian/Kazakh scans. Old Scans category scores only 51.1. | V2 can loop until the token limit, mostly on dense multi-column pages. Key-information extraction scores 21.0 on ThaiOCRBench. |
Sources
- Chandra GitHub README
- Datalab pricing
- GitHub issue #51 token loops
- Typhoon OCR V2 2B model card
- Typhoon OCR technical report
What each one is
Chandra
Datalab's OCR model that keeps layout and handles forms, tables and handwriting.
Chandra 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": "datalab/chandra","fallbacks": ["typhoon/typhoon-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Chandra alternativesTyphoon OCR alternativesBest OCR models in 2026