Apple reportedly plans to use a custom version of Google’s Gemini model to support major Siri upgrades. bloomberg’s Mark Garman. The company could pay Google about $1 billion each year for access to technology that can create summaries and handle planning tasks.
bloomberg Apple says it will continue to rely on its own systems for parts of Siri, while running custom models on its own private cloud computing servers. Gurman reports that the Gemini model uses 1.2 trillion parameters, far more than the 150 billion parameters behind the current cloud-based version of Apple Intelligence.
Apple is preparing to spend about $1 billion annually on a powerful artificial intelligence model built by Google with 1.2 trillion parameters, people familiar with the matter said. bloomberg. The system is expected to play a central role in a major update to Siri, a project the company has been working on for years.
After months of testing, Apple and Google are nearing a deal that would give Apple access to the technology. The people discussing the plan requested anonymity because the talks are private.
Apple has asked Google to rebuild Siri’s core technology, laying the groundwork for a wide range of feature updates planned for next year. The scale of Google’s model will far exceed the AI systems Apple currently uses.
Apple tested other external options, including Google’s Gemini, OpenAI’s ChatGPT, and Anthropic’s Claude, before deciding to move forward with its partnership with Google earlier this year. The goal is to rely on Gemini as a temporary solution until Apple’s own efforts reach the same level.
The updated Siri is scheduled to be released next spring. bloomberg Reported. There are still a few months left until launch, so some plans may change. Apple and Google declined to comment.
Shares of both companies briefly rose after the news surfaced on Wednesday. Apple’s stock rose less than 1% to $271.70, while Alphabet’s stock rose 3.2% to $286.42.
The custom Gemini model is a big leap from the 150 billion parameter system that Apple currently uses in the cloud for Apple Intelligence. This move is aimed at increasing Siri’s ability to handle complex tasks and understand context at a deeper level.
The effort, known internally as Glenwood, is led by Mike Rockwell, creator of the Vision Pro headset, and head of software Craig Federighi. The revamped voice assistant, scheduled to arrive in iOS 26.4, is code-named “Linwood.”
Under the deal, Google’s model will support Siri’s summarizer and planner features, the parts that help the assistant understand information and decide on action steps. Apple’s proprietary model continues to handle some tools and responses.
This model runs on Apple’s private cloud computing servers, and user data is isolated from Google’s systems. Apple has already secured server hardware to support Siri’s new features.
While the partnership is large, Apple has no plans to promote it to consumers. Google acts as a silent technology provider, unlike the visible search contract within Safari. Improvements to Siri are likely to emerge without Google’s branding.
This deal is separate from previous talk of putting Gemini directly within Siri as a chatbot. Those conversations came close to production in 2024 and earlier this year, but never moved forward. Additionally, the new agreement will not place Google AI search functionality within Apple’s operating system and will not change Siri’s search behavior.
During Apple’s latest earnings call, CEO Tim Cook said Siri may have more chatbot options in the future beyond its current ChatGPT options. Apple continues to explore ways to extend Siri without relying on one provider.
Other companies have also adopted Gemini. Snap and several large companies are building products using Google’s Vertex AI platform. For Apple, the move reflects how far behind the company is in the AI space and how willing the company is to use external tools to improve Siri.
Still, Apple has no intention of using Gemini forever. Although the company has lost AI engineers in recent years, including the head of its modeling team, Apple executives still hope to develop its own technology and eventually replace Google’s system within Siri.
Apple’s internal teams are building a proprietary cloud-based model with up to 1 trillion parameters that could be available for consumers as early as next year. This initiative is expected to support Siri’s growth in the long term.
Over time, executives believe they can match Google’s quality. But as Google continues to improve Gemini, it’s becoming harder to close the gap. Its 2.5 Pro version ranks near the top of most large language model comparisons, and this has implications for how Apple plans updates to Siri.
Apple also wants to bring Apple Intelligence and the latest version of Siri to China. Google services are banned in the country, so the systems used there do not rely on Gemini.
Instead, Apple plans to use its own model in conjunction with content filters built by Alibaba Group Holding Ltd. This tool will tailor the response to meet government requirements. Apple is also exploring a partnership with Baidu Inc. for AI capabilities in the Chinese market. bloomberg It was reported earlier this year.
(Photo provided by Omid Armin)
See also: Inside Tim Cook’s efforts to get Apple back into the AI race
Want to learn more about AI and big data from industry leaders? Check out the AI & Big Data Expo in Amsterdam, California, and London. This comprehensive event is part of TechEx and co-located with other major technology events. Click here for more information.
AI News is brought to you by TechForge Media. Learn about other upcoming enterprise technology events and webinars.
