MinerU vs olmOCR
Pick olmOCR if you have an NVIDIA GPU with 12 GB or more and want the higher olmOCR-Bench score (82.4 vs 75.2) for PDF to Markdown at scale. Pick MinerU if you need CPU-only runs or Office files, or you work mostly with Chinese and English papers. The main split is hardware and input range: MinerU runs on CPU and takes many formats, while olmOCR needs a GPU but scores higher.
MinerUopendatalab/mineru | olmOCRallenai/olmocr | |
|---|---|---|
| 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 | Academic PDFs | Bulk PDF to text for datasets |
| Inputs | PDF, images | PDF, images |
| Languages | 80+ languages | Mainly English |
| Max pages per request | 500 | 1,000 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | ✓ Yes |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | ✓ Yes | 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 MinerU if
- It runs on CPU with about 2 GB RAM in the basic tier, so you can start without a GPU.
- It accepts PDF, images, DOCX, PPTX, XLSX, EPUB, OFD and HTML in one tool.
- It can export to Markdown, HTML, LaTeX, DOCX, EPUB and JSON.
Choose olmOCR if
- It scores 82.4 on olmOCR-Bench, above MinerU 2.5.4 at 75.2 in the same table.
- Self-hosting costs under $200 per million pages per Ai2, and the toolkit is Apache-2.0.
- It handles equations, tables and handwriting, and strips headers and footers in reading order.
Key differences
| MinerU | olmOCR | |
|---|---|---|
| Licence | MinerU Open Source License (Apache-2.0 based with extra conditions); MinerU2.5 VLM weights tagged AGPL-3.0 | Apache-2.0 (toolkit) |
| How you run it | Self-host only: CLI, API or WebUI | Self-host with vLLM on a GPU, or use third-party hosted endpoints |
| Hardware | Basic tier: CPU, 2 GB RAM. VLM tiers: 16 GB RAM and a GPU with 8 GB+ VRAM | NVIDIA GPU with at least 12 GB VRAM and 30 GB free disk |
| Price | Free, you pay for compute | Under $200 per million pages self-hosted; DeepInfra hosted at $0.09 input / $0.19 output per million tokens |
| Inputs and outputs | PDF, images, Office files, EPUB, OFD, HTML in; Markdown, HTML, LaTeX, DOCX, EPUB, JSON out | PDFs and images in; Markdown and JSON out |
| olmOCR-Bench | 75.2 (MinerU 2.5.4, independent) | 82.4 (v0.4.0, by the authors) |
| Known failure | Drops merged header rows in tables such as bank statement account lines; repeated content can produce 300 KB figure blocks | Out of memory on a 12 GB GPU; can get stuck on the last few pages; image bounding box coordinates are unreliable |
Sources
- MinerU GitHub README
- olmOCR GitHub README
- MinerU2.5 model card
- Issue #5554 table header rows
- Issue #424 OOM
- Issue #433 stuck
What each one is
MinerU
Open-source PDF to Markdown pipeline popular for academic and Chinese documents.
MinerU 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": "opendatalab/mineru","fallbacks": ["allenai/olmocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
MinerU alternativesolmOCR alternativesBest OCR models in 2026