Jina OCR vs olmOCR
Pick olmOCR for commercial or large-scale PDF to Markdown, because its toolkit is Apache-2.0 and Ai2 puts self-hosting under $200 per million pages. Pick Jina OCR for non-commercial research and evaluation, where its 570M active parameters and hosted API are easy to try. Jina OCR needs a separate license from Jina/Elastic for any commercial use.
Jina OCRjina/jina-ocr | olmOCRallenai/olmocr | |
|---|---|---|
| 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 | 1.5 s / page |
| Best for | Clean Markdown at low cost | Bulk PDF to text for datasets |
| Inputs | PDF, images | PDF, images |
| Languages | About 100 languages | Mainly English |
| Max pages per request | 200 | 1,000 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | ✓ Yes |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | No | No |
| Confidence | No | No |
| Forms | No | No |
| Multilingual | ✓ Yes | No |
| 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
- It has a hosted API as well as open weights, so you can test it without setting up a GPU.
- Only about 570M parameters are active (3B MoE), and it runs at 2.57 pages per second on an A100.
- It scores 91.14 on OmniDocBench v1.6 and 83.4 on olmOCR-Bench (both reported by the authors), and it can output HTML and LaTeX as well as Markdown.
Choose olmOCR if
- The toolkit is Apache-2.0, so you can use it in a commercial product without a separate deal.
- Self-hosting costs under $200 per million pages according to Ai2, and DeepInfra hosts olmOCR-2-7B at $0.09 input / $0.19 output per million tokens.
- It handles equations, tables and handwriting, and strips headers and footers in natural reading order, which suits academic papers and mixed printed and handwritten pages.
Key differences
| Jina OCR | olmOCR | |
|---|---|---|
| Licence | CC-BY-NC-4.0; commercial use needs a license from Jina/Elastic | Apache-2.0 (toolkit) |
| Model size | 3B MoE, about 570M active | 7B |
| Hardware | GPU recommended (tested on A100 and L4); falls back to CPU in Transformers | NVIDIA GPU with at least 12 GB VRAM and 30 GB free disk; not for CPU-only machines |
| Pricing | Free for non-commercial self-hosting (you pay for compute) | Under $200 per million pages self-hosted (Ai2); DeepInfra $0.09 / $0.19 per million tokens |
| Output formats | Markdown, HTML, LaTeX | Markdown, JSON |
| olmOCR-Bench score | 83.4 (authors) | 82.4 on v0.4.0 (authors) |
| Known failure | Hosted API returns HTTP 503 on cold start; retry after 30 to 60 seconds | Out of memory on a 12 GB GPU (issue #424); can get stuck on the last pages (issue #433); unreliable image bounding boxes (issue #437) |
Sources
- jina-ocr-v1 model card
- Jina model page
- olmOCR GitHub README
- olmOCR-2 model
- Issue #424 OOM
- Issue #433 stuck
What each one is
Jina OCR
Compact page-to-Markdown model from Jina AI, built on the DeepSeek-OCR encoder.
Jina OCR detailsolmOCR
Open-source vision model from Ai2 (often searched as olm ocr) that turns PDFs into clean, reading-order text. Latest release: olmOCR 2.
olmOCR 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": ["allenai/olmocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Jina OCR alternativesolmOCR alternativesBest OCR models in 2026