# Datalab Chandra OCR

5 Hacker News commenters recommended Datalab Chandra OCR and 1 criticized it, November 2025 to November 2026. It was recommended in 6 of 13 months.

**Verdict:** People recommend Chandra for its accuracy on text and handwriting, and it is open-weight. Critics note weak layout bounding boxes, slow runtime, and restrictive license terms.

## Why people recommend it

- Accurate on text and handwriting (2 comments): "Chandra is even more accurate on text but its layout bounding boxes are worse and it runs very slowly" ([source](https://news.ycombinator.com/item?id=47370066)) "our new model, chandra, saturates handwriting-heavy benchmarks like this one" ([source](https://news.ycombinator.com/item?id=46315821))

## What people warn about

- Weak layout bounding boxes (2 comments): "Chandra is even more accurate on text but its layout bounding boxes are worse and it runs very slowly" ([source](https://news.ycombinator.com/item?id=47370066)) "otherwise-great OCR models (like Chandra) have terrible bounding boxes" ([source](https://news.ycombinator.com/item?id=48100333))
- Runs very slowly (1 comments): "Chandra is even more accurate on text but its layout bounding boxes are worse and it runs very slowly" ([source](https://news.ycombinator.com/item?id=47370066))
- Restrictive license constraints (1 comments): "which is larger and has bad license constraints" ([source](https://news.ycombinator.com/item?id=46977346))

## Alternatives hackers recommend

- [Mistral OCR](https://hackerapproved.com/p/mistral-ocr.md): 69 recommend
- [PaddleOCR](https://hackerapproved.com/p/paddleocr.md): 19 recommend
- [GLM-OCR](https://hackerapproved.com/p/glm-ocr.md): 16 recommend
- [DeepSeek-OCR](https://hackerapproved.com/p/deepseek-ocr.md): 12 recommend
- [dots.ocr](https://hackerapproved.com/p/dots-ocr.md): 8 recommend

## Cite as

Hacker Approved, "Datalab Chandra OCR", https://hackerapproved.com/p/chandra/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).