Cohere Releases North-Micro-Vision Multimodal Model

CohereLabs · hf · 2026-08-13

CohereLabs has released the North-Micro-Vision-Instruct model on Hugging Face. Built on the Transformers architecture, it focuses on multimodal conversation and image-text-to-text tasks, featuring support for multiple languages and native resolution.

Related event: Cohere Open-Sources North Micro Vision, Its Smallest VLM(7 posts)→

Original post →

More from Models

Models channel →