Apple and Google agree multi‑year pact to power next Siri with Gemini

Apple and Google announced a multi‑year agreement under which Apple will base the next generation of its Apple Foundation Models on Google’s Gemini models and Google Cloud infrastructure, a deal designed to underpin a major overhaul of Siri and expanded Apple Intelligence features later in 2026.

The companies said Apple concluded, "after careful evaluation," that Google’s technology offered the most capable foundation for the models that will sit at the center of forthcoming device and cloud services. The agreement is non‑exclusive, the firms said, and will combine on‑device processing with Apple’s Private Cloud Compute to preserve the privacy commitments that have been central to Apple’s product positioning.

Gemini’s multimodal strengths, handling text, images, audio and other data types, are expected to be integrated into personalization and context‑aware features across Apple’s ecosystem. The rebuilt Siri will draw on those capabilities to provide deeper, more conversational and multimodal assistance, shifting the voice assistant from isolated queries toward a continuous, integrated layer of device intelligence that can better interpret user context and app states.

Financial terms were not disclosed. Industry estimates have suggested the scale of the arrangement could run into the low billions annually, though those figures remain unconfirmed. What is clear is the breadth of the opportunity: Apple’s ecosystem encompasses more than two billion active devices, giving Gemini an immediate addressable market far beyond traditional cloud deployments.

The pact deepens a long‑standing commercial relationship between the two companies and signals a significant vote of confidence in Google’s position among frontier model makers. Observers said the news helped reinforce investor optimism about Google’s AI strategy, contributing to a surge in market enthusiasm that has elevated Alphabet’s market valuation in recent months.

The deal also changes competitive dynamics in the AI landscape. Apple evaluated models from other major suppliers, including OpenAI and Anthropic, before choosing Gemini. Industry analysts have suggested the arrangement will shift how OpenAI is positioned within Apple’s product stack. "It shifts OpenAI into a more supporting role, with ChatGPT remaining positioned for complex, opt‑in queries rather than the default intelligence layer," said Parth Talsania, chief executive of Equisights Research.

Regulatory and legal questions hover over the arrangement. A federal court ruling in late 2025 imposed remedies on some of Google’s prior exclusive distribution deals, and company statements said this agreement does not concern search defaults or Gemini app distribution. Legal experts caution that non‑exclusive language and the separation from search deals will be closely examined by regulators and in public filings as the partnership unfolds.

For Apple, the deal represents a strategic acceleration after months of delays and uneven reception to its earlier AI efforts. Engineers have been experimenting with third‑party models since 2024, and the Google tie‑up offers a rapid path to scale and multimodal sophistication. For Google, access to Apple’s device base widens Gemini’s footprint and entrenches the company further in the core infrastructure of consumer AI.

Integrating large models into billions of devices, while preserving privacy and responsiveness, will be a complex technical undertaking. Apple and Google will spend the coming months refining how on‑device inference, private cloud compute, and Gemini’s capabilities are balanced. The result, expected to arrive later in 2026, could redefine how mainstream users experience AI on their phones, tablets and laptops, and will likely shape regulatory scrutiny and competitive strategy across the industry.