EasyDoc provides a multimodal document processing API designed to convert unstructured documents into structured, hierarchical JSON. This platform is targeted at developers building AI and LLM applications, offering enriched data for model inference, fine-tuning, and optimized context windows by identifying content blocks and extracting semantic information.

How It Works

EasyDoc employs AI to parse documents, moving beyond simple text extraction to identify meaningful content blocks and reconstruct document hierarchies. It supports multimodal content, converting tables, figures, and visual data into machine-readable JSON, enabling LLMs to process complex information more effectively.

Quick Start & Requirements

• Install/Run: Interact via REST API endpoints (POST /api/v1/parse, GET /api/v1/parse/{task_id}/result).
• Prerequisites: Requires an API key obtained from the EasyDoc platform.
• Supported Files: PDF, TXT, DOCX, DOC, PPTX, PPT.
• File Size Limit: 100 MB per file.
• Documentation: API Reference for POST /api/v1/parse, API Reference for GET /api/v1/parse/{task_id}/result.

Highlighted Details

• Content block identification for meaningful data grouping.
• Semantic extraction to reconstruct document hierarchies.
• Multimodal parsing of tables, figures, and visual data into JSON.
• Offers "Lite" and "Pro" modes, with a "Premium" beta.

Maintenance & Community

• Support: Discord community, email: cooperate@easylink-ai.com.
• Roadmap: Not explicitly detailed in the README.

Licensing & Compatibility

• Licensing: Not specified in the README.
• Compatibility: REST API is language-agnostic.

Limitations & Caveats

The "Premium" tier is in beta with a waitlist for API keys. Pricing is per 1000 pages, with different rates for Lite and Pro modes.