Apple to Use Google's Gemini Model for Next-Generation Siri
Key Points
- Apple will integrate a custom version of Google's Gemini model into the next Siri update.
- Gemini will run on Apple’s Private Cloud Compute to handle summarizer and planner functions.
- Apple will continue to use its own models for other Siri tasks, creating a hybrid AI system.
- The deal is reported to involve annual payments of about $1 billion to Google.
- Apple plans to eventually replace Gemini with its own trillion‑parameter cloud model.
Apple is planning to integrate a custom version of Google's Gemini large‑language model into its upcoming Siri update slated for 2026. The partnership will run Gemini on Apple’s Private Cloud Compute infrastructure to handle Siri’s summarizer and planner functions, while the assistant will still rely on Apple‑built models for other tasks. Apple is expected to pay Google roughly $1 billion annually for the technology and aims to eventually replace Gemini with its own trillion‑parameter cloud model for consumer use.
Background of the Siri Upgrade
Apple is preparing a more capable version of its voice assistant, Siri, scheduled for release in 2026. The enhanced assistant is designed to better synthesize information and execute complex tasks on behalf of users, a core element of Apple’s strategy for the updated service.
Integration of Google’s Gemini
According to Bloomberg, Apple will employ a custom‑tuned version of Google’s Gemini large‑language model to power key components of the new Siri. The Gemini model will run on Apple’s Private Cloud Compute servers and will be responsible for Siri’s “summarizer and planner functions,” which help the assistant combine data from multiple sources and decide how to carry out sophisticated requests.
Continued Use of Apple‑Developed Models
While Gemini will handle specific high‑level tasks, Apple will still rely on its own home‑grown models for other aspects of Siri’s operation. This hybrid approach allows Apple to leverage Google’s advanced language capabilities while maintaining a degree of internal control over the assistant’s overall performance.
Financial Terms and Future Plans
Apple is reportedly set to pay Google about $1 billion each year for the use of Gemini. This arrangement is described as a fraction of the amount Google has previously paid Apple to keep Google Search as the default option on Apple devices. Apple’s long‑term goal is to replace Gemini with its own cloud‑based model that is expected to contain roughly one trillion parameters and be ready for consumer applications as early as the following year.
Implications for the AI Landscape
The partnership underscores the growing interdependence among major technology firms in the race to deliver advanced AI‑driven consumer experiences. By integrating Gemini, Apple aims to enhance Siri’s capabilities quickly while it continues to develop its own large‑scale models for future deployment.