NuExtract3: Open-Weight 4B VLM for Markdown, OCR, and Structured Extraction

Core summary

One sentence to understand this update

Numind has released NuExtract3, a self-hostable, open-weight 4B Vision-Language Model (VLM) based on Qwen3.5-4B under Apache-2.0 license, designed for image/text to Markdown, OCR, and structured information extraction.

Impact & opportunity

What this could mean

This open-weight VLM provides developers with a powerful, versatile tool for local information extraction tasks, reducing reliance on proprietary services and fostering custom AI solutions.