NVIDIA surpasses 1,000 public repositories on Hugging Face, including 820 models and 249 datasets

We 💚 open source

LocateAnything is trending top on Hugging Face Models, together with PiD and many awesome models from NVIDIA. A remarkable moment for NVIDIA and teams!

@ClementDelangue @huggingface 💚

@NVIDIAAI Interesting. In the LLM world, 'open source' often refers to public weights. But here's the catch: the majority still come with non-commercial licenses, not fully permissive ones like Apache 2.0. A crucial distinction often missed.

@huggingface Been watching their HF growth for a while, crossing 1k repos is a big deal. The LocateAnything model trending at #1 is well deserved.

@ClementDelangue @huggingface Giving away model weights keeps developers locked inside their proprietary CUDA hardware ecosystem.

That "King of American Open-source AI" framing is a bold one. While their Hugging Face contributions are certainly impressive, it's a stretch when their foundational compute stack (CUDA) remains proprietary. True open-source leadership empowers innovation universally, not just w…

@ClementDelangue @huggingface Nvidia is just Killing it!

just tried their Cosmos 3 Super Model and it's amazing!

@ClementDelangue @huggingface nvidia running the ai meta

@Apreston18gmai1 @huggingface @mslittle1 do you think LocateAnything will actually be useful for production apps or is it mostly research?

@ClementDelangue @huggingface @ClementDelangue 1,000 models on HF is an ecosystem play, not just a model release. I think what NVIDIA figured out early - that developer distribution is the real moat - is exactly what enterprise AI adoption runs on. The chip company became the platform company.

@ClementDelangue @huggingface Interesting metrics! When we talk "King of Open-source AI," does a high volume of models and datasets on HuggingFace truly outweigh the ecosystem-shaping impact of a single foundational model release?

@ClementDelangue @huggingface Wall Street obsession with AI models is getting exhausting.

@NVIDIAAI yes we do! 💚💚

@huggingface @ClementDelangue Bring back the app 😡😡😡

@NVIDIAAI Same me

@NVIDIAAI we 🩷 you

@ClementDelangue @huggingface honnêtement le truc sous-coté ici c'est que l'open-source le plus utile vient de plus en plus de la couche infra, pas des labs. nvidia qui release des world models et des datasets > encore un chat model

@ClementDelangue @huggingface nvidia's really pushing the envelope, but open-source isn't just about numbers. quality and community engagement matter just as much.