Cosmos3-Edge is an open-weight NVIDIA multimodal model distributed through Hugging Face. Developers can download and run it locally or in containers for vision-language and text-generation workloads.
In the Image generation space, Cosmos3 Edge takes a focused approach. It focuses on running an open-weight multimodal vision-language model locally for development and inference. It is built as an open-source project for AI developers and researchers. Cosmos3 Edge is open source under the Open Source license. It runs on the web and the command line, and it can be self-hosted.
Behind Cosmos3 Edge is NVIDIA, and it first shipped in 2026. The project is developed in the open on GitHub with 11.5k stars and 82 commits in the last 90 days. Among its 6 catalogued features are vision-language model, text generation, and image understanding.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do