Chandra

Chandra is an OCR and document intelligence model that converts PDFs, spreadsheets, and slides into structured markdown, HTML, and JSON outputs.

Chandra acts as an OCR engine designed to extract information from complex documents. It converts inputs such as PDFs, spreadsheets, and slides into machine-readable formats like markdown, HTML, and JSON. The model identifies word-level bounding boxes and natural document boundaries for multi-file inputs. Organizations use it to process training-grade corpora, including scientific papers, books, and web crawls. The system supports deployment in managed cloud environments, private VPCs, or air-gapped on-premise servers. It processes over 90 languages and provides extraction tools for audit-ready data.

Chandra at a glance

Best for: Software engineer, Researcher, Finance
Runs on: Web, API, self-hosted
Pricing: Free tier with monthly allowance; Team plan for production; Custom enterprise plans available; pay-as-you-go per processor.
Released: October 2025
Category: AI productivity tool
Website: datalab.to

Key features

Word-level bounding box detection
Support for 90+ languages
Extraction of PDFs, spreadsheets, and slides to structured formats
Identification of document boundaries
Quality evaluation using custom rubrics
Support for on-premise and air-gapped deployments
Pipeline composition and versioning

Use it for

Parsing scientific papers and LaTeX documents
Creating training corpora from long-form books
Processing clinical trial documents and claim forms
Extracting data from complex, multi-page document files

Alternatives to Chandra

Docling — Provides advanced PDF understanding and local execution for sensitive data.
OpenDataLoader PDF — It ranks first in benchmarks for layout and table extraction without needing a GPU.
MiniCPM OCR — Efficient document parsing with correctness assessment and high throughput.
paperless-gpt — Generates AI titles and tags to organize documents automatically.
NuMind — Achieves high accuracy in document extraction and outperforms generalist models.
Xberg — Extracts documents, code, images, and audio using one native-Rust engine.
Paper Xray — Recovers hidden failure modes, author bets, and concrete symbol examples.
MinerU — Parses complex documents and tables into structured markdown for agents.
deepdoctection — Orchestrates layout analysis, OCR, and model fine-tuning in PyTorch pipelines.
docTR — Extracts text and layout details from documents using deep learning.
X-doc AI — Preserves layout and formatting for complex technical files.
Paperless-AI — Integrates AI document analysis and RAG chat into Paperless-ngx workflows.

All Chandra alternatives