Read the text in PDFs and images.
OCR reads the text in PDFs and images, with tables and the position of every block. Save a processor per use, then send it documents from the API, the CLI or the console.
What you get
Text or layout
Get plain text, or JSON layout with the position of each block. Tables are read too.
English and Chinese
The unlimited-ocr model (Baidu Unlimited-OCR 1.0.0, MIT licensed) reads both.
Processors
Save the model, languages and output format once per project, and reuse them by name.
Nothing kept
Documents and results are never stored. Only pages that succeed are charged.
At a glance
- Formats
- PDF, PNG, JPEG, TIFF and WebP
- Per document
- 10 MiB and up to 5 pages
- Default quota
- 10,000 pages per billing period
- API
- ocr.ezghcloud.com
Built on the platform
Governed like every other product, from the first call.
More in ai and machine learning
Inference
Hosted models behind an OpenAI-compatible API
EZGH Inference serves models behind an OpenAI-compatible API. The stock OpenAI SDKs work with only the base URL and API key changed.
MCP Central
Governed access to MCP servers for AI agents
MCP Central gives your organization one endpoint for every MCP server it uses, signed in through EZGH SSO. Admins approve servers and tools; each person picks the ones their clients see.