01 / RECOGNITION
LipiOCR
Turn printed Nepali and English documents into usable text, with automatic language recognition.
DOCUMENT INTELLIGENCE
Your documents.
Ready to work.
Extract text, preserve structure, and connect your platform.
Document→Text & tables→PDF + Markdown
02 / VISUAL CONTENT
Lipi Vision
Work with visual page content and preserve source imagery alongside extracted text.
03 / DOCUMENT STRUCTURE
Lipi Structured
Organize text and tables for PDF and Markdown output. Complex layouts may need review.
LIPIOCR BENCHMARKS
Built for Nepali documents.
Reported LipiOCR evaluation results for printed Nepali text and table extraction.
TEXT RECOGNITION
Printed Nepali
91.2%
Word-level accuracy
LipiOCR · Printed Nepali evaluation
STRUCTURED EXTRACTION
Tables
92
Table extraction score
LipiOCR · OmniDocBench 1.6
Results reported by the Lipi team. Performance varies with document quality and layout.
YOUR WORKSPACE
Documents
BUILD WITH LIPI
API access
Connect your platform
API documents and results are temporary. Only usage records are retained for billing.
Copy this key now. It will not be shown again.
One request. Structured output.
Use a unique Idempotency-Key for each document. Reusing it returns a conflict without another charge. Results are returned as a ZIP containing PDF, Markdown and assets. Keep API keys on your server.
API reference ↗CLEAR, PAGE-BASED PRICING
Usage & billing
50 trial pages per account. Then NPR 2 per successfully processed page. Monthly invoices, manual payment.
OFFICE ADMINISTRATION