NetEase Youdao's Two Open-Source Models Take Top Spots on Hugging Face

Two open-source models released by NetEase Youdao have taken top positions on Hugging Face, according to reports, which also note that developers around the world are accelerating their work to adapt them. NetEase Youdao, the NetEase unit known for education technology and translation tools, has been steadily opening its model capabilities to the open-source community.

The reports do not specify the models' names or exactly which rankings they topped, but the 'double number-one' framing points to popularity or download-related placement on Hugging Face. For open weights, that kind of position usually reflects two things: whether a model is actually being called, and whether the team keeps maintaining the weights and the tooling around them.

The word that matters here is adaptation. Adoption of an open model tends to show up first as compatibility work in inference frameworks, follow-on quantised versions and downstream fine-tuning projects. When developers start building adapters around a model, it has entered the shortlist of models people actually use — not just a name from a launch event.

For Chinese AI teams, open weights are also a distribution channel: once the model is public, developers can verify it in their own environment and only then decide whether it belongs in a production pipeline. That path leaves a deeper trace than a ranking position alone.

What to watch next is whether these two models retain steady usage beyond the leaderboard moment, and whether more detailed model information or follow-up releases appear.