Modelofficial release

nvidia/PixelUMM

nvidia/PixelUMM →Hugging Face ↗
Parameters–
Context–
Input → outputtext + image + audio → text + image + audio
Licenseother
Precision–
Architecture–
ReleasedOct 1, 2026
Downloads0

Based on Qwen/Qwen3-8B

Change history

  1. Oct 18, 2026
    task: image-to-text → any-to-any

This release in the news

  1. Oct 1, 2026 · Model release
    nvidia/PixelUMM