Last Monday, towards the end of the WWDC25 keynote, Craig Federighi said that Apple had “expanded” its vision for Swift Assist, and it would now allow developers to bring their Large Language Model of choice directly to Xcode.

Now, we know more about how this is going to work.

This was a feature that Apple during last year’s WWDC, but hadn’t released until now, on Xcode 26 beta. Initially, as explained by Apple, it was going to serve as:

There’s a new Apple research paper making the rounds, and if you’ve seen the reactions, you’d think it just toppled the entire LLM industry. That is far from true, although it might be the best attempt to bring to the mainstream a discussion that the ML community has been having for ages. Here is why this paper matters.

The paper in question, , is certainly interesting. It systematically probes so-called Large Reasoning Models (LRMs) like Claude 3.7 and DeepSeek-

While Apple has fallen behind the curve in terms of the AI features the company has actually launched, its researchers continue to work at the cutting edge of what’s out there.

In a new paper, they take issue with claims being made about some of the latest AI models – that they are actually capable of step-by-step reasoning. Apple say its tests show that this simply isn’t true …

While it’s acknowledged that conventional generative AI models, aka Large Language Models (LLMs),