GLM-OCR vs dots.ocr
Pick GLM-OCR for high-volume PDF to Markdown, academic papers with formulas and schema-based field extraction, where a $0.03 per million token API or a 0.9B self-hosted model keeps cost low. Pick dots.ocr for multilingual PDFs where you need layout boxes, categories, HTML tables and LaTeX formulas from one self-hosted model. Avoid both for financial tables where every row and decimal must survive: GLM-OCR drops table footer rows and comma decimals, and dots.ocr struggles with complex or very large tables.
GLM-OCRzai/glm-ocr | dots.ocrrednote/dots-ocr | |
|---|---|---|
| Type | Open source | Open source |
| Price | 6 credits / page | 2 credits / page |
| Per 1,000 pages | $6.00 | $2.00 |
| Typical speed | 1.3 s / page | 1.5 s / page |
| Best for | Low-cost general documents | Multilingual document parsing |
| Inputs | PDF, images | PDF, images |
| Languages | 100+ languages | 100+ languages |
| Max pages per request | 500 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | No |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | ✓ Yes | No |
| Forms | No | 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 GLM-OCR if
- You want a hosted option: the Z.ai API costs $0.03 per million tokens for input and output, and you can also self-host with vLLM, SGLang, Ollama or transformers.
- You need to pull fields from forms into a user-defined JSON schema, not just transcribe text.
- You process academic papers or high volumes of PDFs, with a 94.62 OmniDocBench v1.5 score from its model card and 1.86 pages per second on a single replica in the authors' test.
Choose dots.ocr if
- You want one model that returns layout boxes, categories and text, with tables as HTML and formulas as LaTeX.
- You need extra output types such as bounding boxes, or SVG from charts and diagrams through the dots.mocr-svg variant.
- You work with multilingual, mixed table and text business or academic documents and accept self-hosting, where dots.mocr (v1.5) reaches 83.9 on olmOCR-bench.
Key differences
| GLM-OCR | dots.ocr | |
|---|---|---|
| How you run it | Z.ai API, or self-host with vLLM, SGLang, Ollama or transformers | Self-host only, via vLLM or Hugging Face Transformers |
| Cost | API at $0.03 per million tokens for input and output; self-hosted is free apart from compute | Free, self-hosted; you pay for compute |
| Model size | 0.9B | 3B (1.7B LLM backbone) |
| Output formats | Markdown, JSON, plain text | JSON, Markdown, HTML, LaTeX, bounding boxes, SVG |
| Licence | MIT for weights; SDK repo and bundled PP-DocLayoutV3 are Apache-2.0 | MIT, plus a separate dots.ocr license agreement file in the repository |
| Benchmarks (authors' own) | 94.62 on OmniDocBench v1.5; 75.2 on olmOCR-Bench via the API; 29.6 mean on ParseBench (independent) | 79.1 on olmOCR-bench for dots.ocr; 83.9 for dots.mocr (1.5) |
| Known failure modes | Repeats output unless a stop sequence is set; drops table footer and summary rows; loses comma decimal separators in cropped tables; no aarch64 Linux support; API limit of 100 pages, 50 MB per PDF and 10 MB per image | Runs of ellipses or underscores can cause endless repeats; complex tables and formulas stay difficult; a table above 1000 grid cells raised an error and dropped the page in a Docling issue |
Sources
- Hugging Face model card
- Z.ai GLM-OCR docs and pricing
- Issue #240 missing table rows
- dots.ocr GitHub README
- dots.ocr blog / limitations
- Docling issue #5146
What each one is
dots.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": "zai/glm-ocr","fallbacks": ["rednote/dots-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
GLM-OCR alternativesdots.ocr alternativesBest OCR models in 2026