Chandra vs dots.ocr
Pick Chandra if you need multilingual scans, or want a hosted API with Markdown, HTML and JSON output, and your company is under $2M in funding or revenue. Pick dots.ocr if you want MIT-licensed weights you can self-host for free at any company size, and your documents are not form-heavy or table-heavy. On olmOCR-bench the authors' tables put Chandra 2 at 85.8 and dots.ocr 1.5 at 83.9.
Chandradatalab/chandra | dots.ocrrednote/dots-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 | 1.5 s / page |
| Best for | Complex layouts and handwriting | Multilingual document parsing |
| Inputs | PDF, images | PDF, images |
| Languages | 40+ languages | 100+ languages |
| Max pages per request | 500 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | No |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | ✓ Yes | ✓ Yes |
| 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
- Chandra 2 scores 85.8 on olmOCR-bench, above dots.ocr 1.5 at 83.9 in the authors' table.
- It averages 77.8% across 43 languages, so it suits multilingual scanned documents.
- You can skip running GPUs and call the Datalab API at $1, $3 or $7.50 per 1,000 pages, with a $20/month free tier.
Choose dots.ocr if
- The code is MIT-licensed and the model is free to self-host, with no revenue or funding cap to check.
- It returns layout boxes, categories and text in one pass, with tables as HTML and formulas as LaTeX.
- The dots.mocr-svg variant converts charts and diagrams to SVG, which Chandra does not list.
Key differences
| Chandra | dots.ocr | |
|---|---|---|
| Licence | Code Apache-2.0; weights modified OpenRAIL-M. Free for research, personal use and companies under $2M funding or revenue. Paid licence above that. | MIT. The repository also includes a separate dots.ocr licence agreement file. |
| How you run it | Self-host with vLLM or Hugging Face, or use the hosted Datalab API. | Self-host only, with vLLM or Hugging Face Transformers. |
| Price | Self-hosted: free within licence terms. Datalab API: $1 (fast), $3 (balanced) or $7.50 (accurate) per 1,000 pages. | Free, self-hosted. You pay for compute. |
| olmOCR-bench | 85.8 (Chandra 2) | 79.1 (original), 83.9 (dots.mocr / 1.5) |
| Output formats | Markdown, HTML, JSON, bounding boxes | JSON, Markdown, HTML, LaTeX, bounding boxes, SVG |
| Hardware | Single NVIDIA GPU via vLLM. Benchmarked on one H100 80GB. Presets for A100, L40S and RTX 4090/5090. | Single CUDA GPU with vLLM. No VRAM figure published. 3B parameters. |
| Known failure | Token-loop hallucinations reported on 5-6 of about 60 Russian and Kazakh scans. Weak on Urdu (43.2%) and old scans (51.1 on olmOCR-bench). | Runs of ellipses or underscores can make output repeat endlessly. A table above 1000 grid cells raised an error and dropped the page in a Docling issue. |
Sources
- Chandra GitHub README
- Datalab pricing
- GitHub issue #51 token loops
- dots.ocr GitHub README
- dots.ocr blog / limitations
- Docling issue #5146
What each one is
Chandra
Datalab's OCR model that keeps layout and handles forms, tables and handwriting.
Chandra detailsdots.ocr
Single vision model from rednote that does layout detection and content recognition together.
dots.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": ["rednote/dots-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Chandra alternativesdots.ocr alternativesBest OCR models in 2026