Bringing AI to the Browser – On-Device AI with the Web AI Toolkit

In the current rush to adopt AI, much of the focus is on using large language models hosted by the likes of OpenAI, Anthropic and Google. But there’s a different approach which won’t solve for all the cases those large models do, but is increasingly capable. It’s the approach Apple has taken with Apple Intelligence–running inference on the device where that makes sense, with smaller more specialised models and only relying on larger, cloud base solutions where the local results won’t suffice.

This is something that is increasingly coming to the Web, with a number of potential solutions, including the Web AI Toolkit which you can learn more about here.