Reports On May 6, 2026 say Apple's next Siri will be powered by Google's Gemini models and will allow iPhone users

Reports on May 6, 2026 say Apple’s next Siri will be powered by Google’s Gemini models and will allow iPhone users to select third‑party AI models and associated custom voices (examples cited include Google and Anthropic).

On May 6, 2026 reporting summarized earlier confirmations and follow‑up details that Apple’s forthcoming Siri will use Google’s Gemini models and let users pick third‑party AI chatbot models to handle responses. Earlier coverage last month first described third‑party integration; the newer reporting adds specifics on how model selection and voice options tied to external providers could work.

Gemini’s Personal Intelligence capability, already demonstrated in Google’s ecosystem, can extract details from text, photos and videos across a wide set of apps-including Gmail, Calendar, Drive, Google Photos, YouTube watch history and several search services (Search, Shopping, News, Maps, Flights, Hotels) — to personalize replies. Apple’s implementation is expected to mirror that app‑level access for Apple services such as Mail, Calendar, Photos and Notes, enabling assistant responses that draw on a user’s personal content.

A planned Extensions system would let third‑party model apps plug into Apple’s assistant framework so users can choose which model answers queries inside Siri, Writing Tools and other features. The reporting gives concrete examples: third‑party apps for Gemini and for Anthropic’s Claude could add Extensions support, allowing those models — and the custom voices they offer — to respond whenever a user selects them.

This shift is significant because Apple first began advertising expanded Siri capabilities in 2024 but has repeatedly delayed broad availability. Moving from promotional descriptions to explicit integrations and demonstrated app‑level intelligence signals a more concrete product direction and raises expectations that these capabilities could reach users sooner than an all in‑house development path would allow.

The coverage highlights three operational implications. First, enabling external models to plug into Siri reduces Apple’s dependence on completing its own large‑model development before shipping advanced assistant features. Second, giving users choice preserves continuity for people who already rely on particular chatbots or custom voices. Third, opening the platform to multiple providers is likely to accelerate competition and improvements among the available models.

For developers and platform teams, the immediate takeaway is both technical and strategic: prepare for an Extensions architecture that routes assistant requests to multiple backends, support model‑specific voice and UI cues, and plan for differing privacy and data‑access semantics across providers. Apple may still pursue its own models, but this path lets third‑party and first‑party providers iterate in parallel while users gain earlier access to more capable Siri features.