ClawTrust
2.7k Downloads
Overview
High-accuracy cloud OCR and document extraction via Nanonets/DocStrange, converting PDFs and images into structured Markdown, JSON, or CSV with optional confidence scores and layout metadata.
Key Advantages
1.Supports multiple output formats (markdown, JSON, CSV) in a single or combined request, covering raw text, fields, and tables.
2.Rich structured extraction options including JSON field lists, JSON Schema-based typing, table-to-CSV, hierarchy output, and bounding boxes.
3.Per-field confidence scoring (0–100) to support downstream validation, QA, and human-in-the-loop review pipelines.
4.Async and sync endpoints with clear guidance by document size, helping avoid timeouts on larger documents.
5.Security-conscious documentation with strong emphasis on secret management, privacy risks, and operational safeguards for API keys and sensitive files.
Use Cases
- Automated invoice and receipt processing, including totals, vendors, dates, and line items with confidence scores.
- Contract and legal document text extraction to Markdown for downstream analysis or summarization by other tools/agents.
- Bank statement and other financial document parsing, especially with financial-docs markdown options and table extraction.
- General OCR for scanned PDFs and images to obtain clean text or structured JSON for indexing, search, or analytics.
- Form digitization where specific fields are requested via JSON field lists or JSON Schema for structured capture.
Evaluation Scores
8.3
/ 10
Reliability
7.5
Functionality
9.0
Usability
9.0
Safety
7.5
Performance
8.0
Compatibility
8.5
Based on 1 evaluation · Latest: 3/19/2026
Download Trend
Loading...
Evaluation History (1)
8.3/103/19/2026▼
OS: darwin-x64LLM: minimax/minimax-m2.5
DocStrange (Nanonets OCR) is a capable cloud OCR and document-extraction skill that turns PDFs/images into Markdown, JSON, or CSV, with strong support for structured fields, tables, and per-field confidence scores. It’s well-suited for invoice/receipt processing, financial documents, contracts, and general document OCR where you need structured outputs and layout-aware metadata. The main risks are data privacy and dependence on an external SaaS: documents are sent to Nanonets’ servers, so you must be comfortable with their security and retention practices and avoid highly sensitive data until vetted. Configuration and usage are straightforward (env-based API key, clear sync/async patterns, good examples and troubleshooting), making it a strong choice whenever your agent needs reliable document OCR and extraction and you accept third-party processing of the documents.
Comments (0)
No comments yet. Be the first!