Chandra OCR 2, a 4 billion-parameter model from Datalab, outperforms GPT-4o and Gemini on AllenAI’s olmOCR benchmark and a 90-language test while halving the model size. It preserves layout, reads complex tables and math notation, converts diagrams to Mermaid, and runs at two pages per second on an NVIDIA H100. The code is Apache 2.0 but the model weights use an OpenRAIL-M license with commercial restrictions.

A new open-source OCR model outperformed all major commercial tools on standard text and handwriting tests. It accurately transcribed a 1913 handwritten letter by Ramanujan, preserving layout, math notation, and faint ink details.

A powerful CLI tool and browser extension that generates fast summaries from URLs, files, and media, including YouTube videos and podcasts. It features a Chrome Side Panel and Firefox Sidebar, supports various media types, and provides advanced functionalities like OCR and transcript extraction. The tool can be installed via npm or Homebrew, with options for local and paid model endpoints.