Apple is developing a new Siri chatbot to compete with ChatGPT, Gemini, and Claude, with a debut planned for iOS 27 at WWDC. According to Bloomberg, the integrated assistant will search the web, generate content, and analyze files directly within Apple's ecosystem. The major unanswered question is pricing, given Apple's history of bundling core services for free.

Running a large language model across billions of devices requires massive server infrastructure and electricity. Apple is reportedly considering using Google's servers for its custom Gemini-based model, which would incur significant costs. This mirrors the expenses faced by OpenAI and Google, who spend billions annually on inference. Apple must decide whether to absorb these costs or pass them to users.

A likely model follows iCloud's tiered approach: a basic, throttled version free for all, with an advanced subscription for full functionality. This would align with the industry's $20/month standard set by competitors. Launching as a free service could rapidly attract users away from paid rivals, but sustaining the infrastructure long-term may necessitate a paid tier. The outcome will shape Apple's AI strategy for years.