Apple may be delivering on its promise to add more AI models to Apple Intelligence. The company recently rolled out a backend update with iOS 18.4 that includes coding language for supporting Google Gemini. MacRumors analyst Aaron Perris initially spotted the update on X.
The report suggests Apple may be working to integrate the technology into its next software update, but a launch date for Gemini support remains uncertain. Apple executive Craig Federighi previously hinted that Gemini would find its way onto the iPhone during Apple’s Worldwide Developers Conference in June 2024. A representative for Apple did not respond to a request for comment.Â
Apple’s desire to add more AI models to Apple Intelligence could make for a more dynamic user experience. It may potentially lean on certain models for some tasks, and other models for separate tasks.
Read more: What Is Gemini? Everything You Should Know About Google’s AI Too
It also comes as Apple intends to make Siri a more personalized, smarter virtual assistant amid a broader push to strengthen the company’s presence in smart home technology, where it has traditionally lagged behind Google and Amazon.
At the same time, Apple is expected to launch a generative AI-powered version of Siri in the weeks ahead, following reports of glitches and delays. Apple Intelligence currently allows users to ask Siri questions that are processed by OpenAI’s ChatGPT.
But some previously announced features have not yet been introduced, such as the ability to better respond to tasks, offer answers with greater personalized insight and do more within apps.
Read more:Â Best iPhone in 2025: Here’s Which Apple Phone You Should Buy
Meanwhile, Google has recently introduced updates for Gemini, including the ability to recall past chats and multimodal capabilities that support text, images and automated tasks.
Reece Hayden, an analyst at ABI Research, said the approach to include multiple AI models makes sense for a few reasons, particularly because varying models offer different accuracy, data and performance. “Deploying multiple [options] means that applications can choose the most appropriate model to perform a task and optimize cost and performance,” Hayden told CNET.
He added that by using multiple models, Apple can better keep customer data safely within the Apple Intelligence ecosystem instead of handing it over to third-party systems. “Sole partnership with OpenAI is hugely risky,” he said. “By expanding the ecosystem, they gain additional buyer power which can support contract negotiations in the future.”