News for hackers. Stuff that matters.
Re:DocStrange: An LLM-Ready Data Platform That Performs Bett... by souvik3333 on Wednesday October 15, 2025 @08:13AM (#45591189)
We have developed DocStrange to create LLM-ready data from images and PDFs. We have open-sourced a 3B finetuned model also. You can try both the open-sourced and private models from the demo.

HF: https://huggingface.co/nanonets/Nanonets-OCR2-3B Demo: https://docstrange.nanonets.com/ Blog: https://nanonets.com/research/nanonets-ocr-2/

This model is an improvement over our last open-source model. We have fixed some of the issues that the community faced and some of the features that were requested (handwritten, multi-lingual).

The models are trained on 3 million documents, including handwritten documents, financial reports, complex tables, documents with watermarks, and stamps. Feel free to try it and share feedback.

Re:DocStrange: An LLM-Ready Data Platform That Performs Bett... by AdityaNahata on Wednesday October 15, 2025 @08:14AM (#45591205)
Do you guys provide api support also? I am processing documents for a project
Re:DocStrange: An LLM-Ready Data Platform That Performs Bett... by souvik3333 on Wednesday October 15, 2025 @08:15AM (#45591216)
Yeah, we do have api support. Currently, you can process 10k documents per month free. Let me know if you face any issues.
Re:DocStrange: An LLM-Ready Data Platform That Performs Bett... by ashish2091 on Wednesday October 15, 2025 @08:20AM (#45591276)
Is it better than gemini pro or flash? do you have any benchmarking data? I want to use it for markdown from scanned pdfs.
Re:DocStrange: An LLM-Ready Data Platform That Performs Bett... by souvik3333 on Wednesday October 15, 2025 @08:23AM (#45591301)
We have evaluated against Gemini-2.5-flash. You can check the benchmarks here https://nanonets.com/research/nanonets-ocr-2/#markdown-evalu...