Docling vs olmOCR
Pick Docling to ingest born-digital PDFs and Office files into a RAG pipeline that needs structured JSON with layout. Pick olmOCR to turn scanned or mixed printed and handwritten PDFs into Markdown at volume, if you have an NVIDIA GPU with 12 GB or more. The main split is scope: Docling is a multi-format parser that relies on a plugged-in OCR engine for scans, while olmOCR is a 7B vision-language model built for page-to-Markdown conversion.
Doclingibm/docling | olmOCRallenai/olmocr | |
|---|---|---|
| Type | Open source | Open source |
| Price | 2 credits / page | 2 credits / page |
| Per 1,000 pages | $2.00 | $2.00 |
| Typical speed | 2.2 s / page | 1.5 s / page |
| Best for | Structured export for RAG | Bulk PDF to text for datasets |
| Inputs | PDF, images, DOCX | PDF, images |
| Languages | Latin scripts, CJK | 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 | No | 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 Docling if
- You need one API for PDF, DOCX, PPTX, XLSX, HTML, EPUB, images, emails and XBRL.
- You want lossless JSON export that keeps layout, reading order and table structure, plus Markdown, HTML, DocTags and plain text.
- You have no GPU: it runs on CPU on macOS, Linux and Windows (x86_64 and arm64), and an MIT licence covers the library.
Choose olmOCR if
- You process large batches of PDFs or images into Markdown, and Ai2 puts self-hosted cost under $200 per million pages.
- Your pages have equations, tables or handwriting, and you want headers and footers stripped in natural reading order.
- You want a published score: 82.4 on olmOCR-Bench, ahead of Marker (76.1), MinerU 2.5.4 (75.2) and Mistral OCR (72.0) in the same table.
Key differences
| Docling | olmOCR | |
|---|---|---|
| What it is | Python library from IBM Research that converts many file formats into a structured DoclingDocument | 7B vision-language model and toolkit from Ai2 that converts PDFs and images to Markdown |
| How you run it | Local library; CPU works, GPU speeds up model stages | Self-host with vLLM on an NVIDIA GPU (12 GB VRAM minimum, 30 GB free disk) or use third-party hosted endpoints |
| Licence | MIT (individual models have their own licences) | Apache-2.0 (toolkit) |
| Price | Free, self-hosted; you pay for compute | Under $200 per million pages self-hosted, per Ai2; DeepInfra hosts olmOCR-2-7B at $0.09 input / $0.19 output per million tokens |
| Output formats | Markdown, HTML, JSON, DocTags, plain text | Markdown, JSON |
| Benchmark | Not published | 82.4 on olmOCR-Bench (v0.4.0), reported by the authors |
| Known failure | Version 2.143.1 drops words and numbers with full-page Tesseract OCR (issue #5254); neighbouring columns can merge on some layouts (issue #5132) | Runs out of memory on a 12 GB GPU (issue #424); can get stuck on the last few pages (issue #433); image bounding box coordinates are unreliable (issue #437) |
Sources
- Docling GitHub README
- Supported formats
- Issue #5254 OCR regression
- olmOCR GitHub README
- Issue #424 OOM
- Issue #437 bbox
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 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": "ibm/docling","fallbacks": ["allenai/olmocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Docling alternativesolmOCR alternativesBest OCR models in 2026