Feb 15, 2026
A natively multimodal model family combines vision, reasoning and agent tasks.
Downloadable model weights. License terms and commercial permissions vary by release; this is not a claim that the entire training stack is open.
Explore on Humanity Unfolds
