OCRRouter

GLM-OCR vs MinerU

Pick GLM-OCR for high-volume PDF to Markdown through a cheap API, or when you want a small 0.9B model under the MIT licence. Pick MinerU when you need to self-host, parse Office files as well as PDFs, or run on CPU with no GPU. On the independent ParseBench, MinerU2.5 scores 45.9 against 29.6 for GLM-OCR.

GLM-OCRzai/glm-ocrMinerUopendatalab/mineru
TypeOpen sourceOpen source
Price6 credits / page2 credits / page
Per 1,000 pages$6.00$2.00
Typical speed1.3 s / page2.0 s / page
Best forLow-cost general documentsAcademic PDFs
InputsPDF, imagesPDF, images
Languages100+ languages80+ languages
Max pages per request500500
Tables✓ Yes✓ Yes
HandwritingNoNo
Math✓ Yes✓ Yes
Bounding boxes✓ Yes✓ Yes
Confidence✓ YesNo
FormsNoNo
Multilingual✓ Yes✓ Yes
On OCRRouterAvailableAvailable

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

  • The Z.ai API costs $0.03 per million tokens for input and output, and you can also self-host the 0.9B model for free.
  • The model weights are MIT licensed, with Apache-2.0 for the SDK repo and the bundled layout model, so commercial use is simple.
  • It can extract fields into a user-defined JSON schema, which suits forms, and it scores 94.62 on OmniDocBench v1.5 per its model card.

Choose MinerU if

  • It reads PDF, images, DOCX, PPTX, XLSX, EPUB, OFD and HTML, and writes Markdown, HTML, LaTeX, DOCX, EPUB or JSON.
  • The basic ONNX tier runs on CPU with 2 GB RAM, so you can start without a GPU.
  • It scores 45.9 on ParseBench against 29.6 for GLM-OCR, and the VLM tier reaches 2.12 pages per second on one A100 with vLLM.

Key differences

GLM-OCRMinerU
LicenceMIT for model weights; Apache-2.0 for the SDK repo and PP-DocLayoutV3MinerU Open Source License (Apache-2.0 based with extra conditions); MinerU2.5 VLM weights tagged AGPL-3.0
How you run itZ.ai API, or self-host with vLLM, SGLang, Ollama or transformersSelf-host only, as a CLI, API or WebUI
PriceAPI: $0.03 per million tokens for input and output; self-hosted is free apart from computeFree, self-hosted; you pay for compute
Input and outputOutput is Markdown, JSON or plain text; the API caps input at 100 pages, 50 MB per PDF and 10 MB per imageTakes PDF, images and Office files; outputs Markdown, HTML, LaTeX, DOCX, EPUB or JSON
BenchmarksOmniDocBench v1.5 94.62 and olmOCR-Bench 75.2 (authors); ParseBench mean 29.6 (independent)olmOCR-Bench 75.2 and ParseBench mean 45.9 (both independent); OmniDocBench Not published
HardwareCannot run on aarch64 Linux; model size 0.9BBasic tier on CPU with 2 GB RAM; VLM tiers need 16 GB RAM and a GPU with 8 GB+ VRAM; VLM model is 1.2B
Known failuresRepeated output unless a stop sequence is set; drops table footer and summary rows; loses comma decimal separators in cropped tablesVLM pipeline can repeat content, with 300 KB figure blocks; drops merged header rows in tables such as bank statement account lines; converts full-width Chinese punctuation to half-width with no switch

Sources

What each one is

GLM-OCR

Small open-source OCR model from Z.ai with layout regions. MIT licensed.

GLM-OCR details

MinerU

Open-source PDF to Markdown pipeline popular for academic and Chinese documents.

MinerU details

Try both on your own file

Run the file once with each model and compare the output.

PDF, PNG, JPG, WebP, HEIC. Up to 20 MB each, 10 files.No file?

Use one, fall back to the other

{"model": "zai/glm-ocr","fallbacks": ["opendatalab/mineru"],"file_url": "https://example.com/document.pdf"}

Questions

More comparisons

GLM-OCR alternativesMinerU alternativesBest OCR models in 2026