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
• Math equation support (LaTeX)
• Precise table reconstruction
• Support for 40+ languages
• Two inference modes: local (HuggingFace) and remote (vLLM server)

Benchmarks

Model	ArXiv	Old Scans Math	Tables	Old Scans	Headers and Footers	Multi column	Long tiny text	Base	Overall
Datalab Chandra v0.1.0	81.4	80.3	89.4	50.0	88.3	81.0	91.6	99.9	82.7 ± 0.9
Datalab Marker v1.10.0	83.8	69.7	74.8	32.3	86.6	79.4	85.7	99.6	76.5 ± 1.0
Mistral OCR API	77.2	67.5	60.6	29.3	93.6	71.3	77.1	99.4	72.0 ± 1.1
Deepseek OCR	75.2	67.9	79.1	32.9	96.1	66.3	78.5	97.7	74.2 ± 1.0
Nanonets OCR	67.0	68.6	77.7	39.5	40.7	69.9	53.4	99.3	64.5 ± 1.1
GPT-4o (Anchored)	53.5	74.5	70.0	40.7	93.8	69.3	60.6	96.8	69.9 ± 1.1
Gemini Flash 2 (Anchored)	54.5	56.1	72.1	34.2	64.7	61.5	71.5	95.6	63.8 ± 1.2
Qwen 2.5 VL (No Anchor)	63.1	65.7	67.3	38.6	73.6	68.3	49.1	98.3	65.5 ± 1.2
Qwen 3 VL	70.2	75.1	45.6	37.5	89.1	62.1	43.0	94.3	64.6 ± 1.1
olmOCR v0.3.0	78.6	79.9	72.9	43.9	95.1	77.3	81.2	98.9	78.5 ± 1.1

Installation

From PyPI (Recommended)

pip install chandra-ocr

From Source

git clone https://github.com/yourusername/chandra.git
cd chandra
uv sync
source .venv/bin/activate

Usage

CLI

Process single files or entire directories:

# Single file, with vllm server (see below for how to launch)
chandra input.pdf ./output --method vllm

# Process all files in a directory with local model
chandra ./documents ./output --method hf

# Process specific pages with custom settings
chandra document.pdf ./output --page-range "1-10,15,20-25" --max-workers 8

CLI Options:

• --method [hf|vllm]: Inference method (default: vllm)
• --page-range TEXT: Page range for PDFs (e.g., "1-5,7,9-12")
• --max-output-tokens INTEGER: Max tokens per page
• --max-workers INTEGER: Parallel workers for vLLM
• --include-images/--no-images: Extract and save images (default: include)
• --include-headers-footers/--no-headers-footers: Include page headers/footers (default: exclude)
• --batch-size INTEGER: Pages per batch (default: 1)

Output Structure:

Each processed file creates a subdirectory with:

• <filename>.md - Markdown output
• <filename>.html - HTML output
• <filename>_metadata.json - Metadata (page info, token count, etc.)
• images/ - Extracted images from the document

Streamlit Web App

Launch the interactive demo for single-page processing:

chandra_app

The web interface allows you to:

• Upload PDFs or images
• Select specific pages from PDFs
• View OCR results with layout visualization
• Download markdown output
• See extracted images embedded in the output

Inference Modes:

• hf: Loads model locally using HuggingFace Transformers (requires GPU)
• vllm: Connects to a running vLLM server for optimized batch inference

vLLM Server (Optional)

For production deployments or batch processing, use the vLLM server:

python scripts/start_vllm.py

This launches a Docker container with optimized inference settings. Configure via environment variables:

• VLLM_API_BASE: Server URL (default: http://localhost:8000/v1)
• VLLM_MODEL_NAME: Model name for the server (default: chandra)
• VLLM_GPUS: GPU device IDs (default: 0)

Configuration

Settings can be configured via environment variables or a local.env file:

# Model settings
MODEL_CHECKPOINT=datalab-to/chandra-0.2.8
MAX_OUTPUT_TOKENS=8192

# vLLM settings
VLLM_API_BASE=http://localhost:8000/v1
VLLM_MODEL_NAME=chandra
VLLM_GPUS=0

Output Formats

Chandra provides three output formats:

1. HTML: Structured HTML with layout blocks and bounding boxes
2. Markdown: Clean, readable Markdown conversion
3. Layout Image: Visual representation of detected layout blocks