Reports say a Gemini model powers the new Siri. Our read

Updated June 11, 2026 · ~2 min read · Ilura Technology

Source: TechCrunch

DIRECT ANSWERAccording to reporting by TechCrunch, CNBC, and Bloomberg, the rebuilt Siri layer runs partly on a custom Google Gemini model inside Apple's Private Cloud Compute; Apple's own release does not name Gemini. Private Cloud Compute is serious privacy engineering, but hard requests still leave the device. On-device writing keeps a simpler promise: the text never leaves at all.

What was announced?

Apple unveiled the rebuilt Siri at WWDC26 on June 8. What Apple did not say on stage is whose model powers it. According to reporting by TechCrunch, CNBC, and Bloomberg, the new Siri layer runs partly on a custom Google Gemini model — reportedly around 1.2 trillion parameters, at a reported cost of about one billion dollars per year — hosted inside Apple’s Private Cloud Compute. Apple’s own press release does not mention Gemini. Until Apple or Google says it outright, treat the specifics as well-sourced reporting, not confirmation.

What does it mean for personal, on-device writing?

Two readings, and both can be true. The first is about economics. Building a frontier model is now so expensive that even Apple, the most vertically integrated company in tech, reportedly rents one for its assistant’s hard cases. The cloud race has a price of admission few can pay.

The second reading is the one that touches your writing. Running a Gemini-derived model inside Private Cloud Compute is serious privacy engineering. Requests are processed in Apple’s hardened cloud environment rather than shipped off to a general-purpose third-party cloud. That deserves credit, and it is a meaningfully better design than ordinary cloud AI.

But the line that matters stays where it was. When a request is too heavy for the device, your text leaves the phone. Protected in transit, processed in a controlled environment — and still, for a moment, out of your hands. “Strongly protected once it leaves” and “never leaves” are different promises. Most days, for most messages, the difference is academic. For some messages — health, money, a conflict at work — it is the entire point.

Where does ILURA stand?

This is a contrast, not an attack. Private Cloud Compute deserves the credit it gets, and the reported deal mostly shows how costly frontier cloud writing has become. ILURA simply lives on the other side of the line. Drafts are generated on-device with Apple Intelligence. There is no account, and the App Store privacy label reads Data Not Collected. Your text never leaves the device because there is nowhere for it to go. Next to Siri’s cloud tier, that remains a real and honest differentiator.

Message → rule → agent

Read the signal through ILURA

Platform news matters when it changes what users expect from personal AI. ILURA reads these shifts through one lens: private agents trained by the user on iPhone.

Try it now

Put this to work on a real message.

Open ILURA, bring in a message you actually need to handle today, and get it done in your voice — free, on device, no account. It learns the preference, so the behavior carries to the next one.

Free to start · No account · Data Not Collected

Quick answers

Has Apple confirmed that Siri uses Gemini?
No. Apple's own announcement does not mention Gemini. The detail comes from reporting by TechCrunch, CNBC, and Bloomberg, including the reported model size and yearly cost.
Is Private Cloud Compute the same as on-device AI?
No. Private Cloud Compute handles requests on Apple-controlled servers when a task is too heavy for the device. On-device AI means the text never leaves your phone at all.

Related

ILURA does this on your iPhone — on device, private. Get ILURA — free, no account