I’ve been working on the Web Models API, a proposal that lets websites request specific open-weight models and run them on the user’s device. The browser manages the model download, storage and execution, while users control which sites can access each model.
The goal is to give developers a way to use the models they’ve built and tested against without managing the inference runtime in their application. Model resources can be shared across sites, and prompts don’t need to be sent to an AI provider.
The proposal uses a registry of supported models, with model IDs tied to specific releases. If a requested model isn’t supported, the request fails rather than silently substituting another. The API covers generation, streaming, embeddings and sessions.
I’ve built an early implementation in Common, a Chromium-based browser currently in private alpha. The explainer and API reference are open for review; the browser implementation isn’t public yet.
I’d be interested in feedback from people building models and applications: