Chandra

2 likes

Chandra is a highly accurate OCR model that converts images and PDFs into structured HTML/Markdown/JSON while preserving layout information.

Chandra screenshot 1

Cost / License

Free
Open Source (Apache-2.0)

Origin

United States

Platforms

Mac
Windows
Linux
Python

Chandra alternatives

2likes

0articles

Features

OCR

Tags

Chandra News & Activities

Highlights All activities

Recent activities

POX added Chandra as alternative to PDFTrail
23 days ago
Danilo_Venom added Chandra as alternative to DocSnap, Zettelrobbe, Paperless AI and Texify.it + 4 similar activities
3 months ago
eliasbuenosdias liked Chandra
6 months ago
lukas-hecker added Chandra as alternative to Inksight
7 months ago
niksavc liked Chandra
10 months ago
K0RR added Chandra
10 months ago
POX updated Chandra
10 months ago
K0RR added Chandra as alternative to OSS Document Scanner, ABBYY FineReader PDF, Open Scanner and Scan Thing: Scan Anything + 23 similar activities
10 months ago

Chandra information

Developed by
Datalab
Licensing
Open Source (Apache-2.0) and Free product.
Written in
Python
Alternatives
28 alternatives listed
Badge
Get an embeddable badge for Chandra
Supported Languages
- English

AlternativeTo Categories

AI Tools & Services, Office & Productivity

GitHub repository

12,262 Stars
1,244 Forks
61 Open Issues
Updated Jun 26, 2026

Popular alternatives

Chandra was added to AlternativeTo by K0RR on Nov 10, 2025 and this page was last updated Nov 10, 2025.

No comments or reviews, maybe you want to be first?

Official Links

AppStores & Other Links

Social Networks

What is Chandra?

Chandra is a highly accurate OCR model that converts images and PDFs into structured HTML/Markdown/JSON while preserving layout information.

Features:

Convert documents to markdown, html, or json with detailed layout information
Good handwriting support
Reconstructs forms accurately, including checkboxes
Good support for tables, math, and complex layouts
Extracts images and diagrams, with captions and structured data
Support for 40+ languages
Two inference modes: local (HuggingFace) and remote (vLLM server)