olmOCR vs Chandra
Pick Chandra for multilingual scans, HTML or JSON output with bounding boxes, or a hosted API at $1 to $7.50 per 1,000 pages. Pick olmOCR for large-scale PDF to Markdown on your own GPU under a plain Apache-2.0 toolkit licence. Chandra 2 scores higher on olmOCR-bench (85.8 vs 82.4), but its weights carry revenue limits for commercial self-hosting.
olmOCRallenai/olmocr | Chandradatalab/chandra | |
|---|---|---|
| Type | Open source | Open source |
| Price | 2 credits / page | 2 credits / page |
| Per 1,000 pages | $2.00 | $2.00 |
| Typical speed | 1.5 s / page | 2.0 s / page |
| Best for | Bulk PDF to text for datasets | Complex layouts and handwriting |
| Inputs | PDF, images | PDF, images |
| Languages | Mainly English | 40+ languages |
| Max pages per request | 1,000 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | ✓ Yes |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | No | ✓ Yes |
| Confidence | No | No |
| Forms | No | ✓ Yes |
| Multilingual | No | ✓ 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 olmOCR if
- Its toolkit is Apache-2.0 and the weights, training data and code are all released, so there is no revenue cap to track.
- Self-hosting costs under $200 per million pages, per Ai2, which suits bulk PDF to Markdown for LLM training or RAG.
- It handles equations, tables and handwriting, and strips headers and footers in natural reading order, which fits academic papers and mixed printed and handwritten pages.
Choose Chandra if
- It scores 85.8 on olmOCR-bench (Chandra 2) against 82.4 for olmOCR 2, and averages 77.8% across 43 languages.
- It returns Markdown, HTML and JSON plus layout chunks with bounding boxes and labels per block, while olmOCR's image coordinates are unreliable.
- You can skip the GPU and call the hosted Datalab API at $1, $3 or $7.50 per 1,000 pages, with a free tier of $20 per month.
Key differences
| olmOCR | Chandra | |
|---|---|---|
| Licence | Apache-2.0 (toolkit); weights, training data and code released | Code Apache-2.0; weights under modified OpenRAIL-M, free for research, personal use and companies under $2M funding or revenue |
| How you run it | Self-host with vLLM on an NVIDIA GPU with at least 12 GB VRAM, or use third-party hosted endpoints | Self-host on a single NVIDIA GPU via vLLM or Hugging Face, or use the hosted Datalab API |
| Price | Under $200 per million pages self-hosted; DeepInfra $0.09 input / $0.19 output per million tokens | Free self-hosted within licence terms; Datalab API $1 (fast), $3 (balanced) or $7.50 (accurate) per 1,000 pages |
| Output formats | Markdown, JSON | Markdown, HTML, JSON, bounding boxes |
| olmOCR-bench overall | 82.4 | 85.8 (Chandra 2) |
| Languages | Not published | 77.8% average across 43 languages; 43.2% on Urdu and 51.1% on Sanskrit |
| Known failure | Out of memory on a 12 GB GPU (issue #424); gets stuck on the last few pages (issue #433) | Token-loop hallucinations and missing pages on some Russian and Kazakh scans (issue #51); Old Scans category only 51.1 on olmOCR-bench |
Sources
- olmOCR GitHub README
- Issue #424 OOM
- Issue #437 bbox
- Chandra GitHub README
- Datalab pricing
- GitHub issue #51 token loops
What each one is
olmOCR
Open-source vision model from Ai2 (often searched as olm ocr) that turns PDFs into clean, reading-order text. Latest release: olmOCR 2.
olmOCR detailsChandra
Datalab's OCR model that keeps layout and handles forms, tables and handwriting.
Chandra 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": "allenai/olmocr","fallbacks": ["datalab/chandra"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
olmOCR alternativesChandra alternativesBest OCR models in 2026