After Apple introduced a multi-model AI strategy for Siri, Samsung has also begun to follow a similar path and plans to integrate Perplexity's AI technology into its own Bixby voice assistant in the upcoming Galaxy S26 series.

According to well-known whistleblower @chunvn8888 on social platform

Currently, Apple's Apple Intelligence has used local models to handle simple tasks, while more complex reasoning and generation tasks are completed with the help of OpenAI's ChatGPT. The combination of Samsung and Perplexity follows a similar division of labor logic. However, Samsung and Google have already established a deep AI cooperation, and the latest Galaxy devices have Gemini-powered AI capabilities built into their customized One UI interface. The introduction of Perplexity this time shows that Samsung is trying to expand its AI model camp instead of relying solely on the Google ecosystem.

fce3de79-94a2-400a-badb-9894bd0dd3f5.png

At the same time, Apple is also actively expanding its multi-model AI strategy. According to Bloomberg, Apple is developing a new generation of Siri with more powerful capabilities. Advanced functions will rely on Google Gemini (such as summarization and multi-step planning), while some functions are still supported by Apple’s self-developed models. The new version of Siri was originally planned to debut with iOS 18, but due to Apple's decision to reconstruct the core architecture of the assistant, it is initially expected to be launched in a software update early next year.

It is reported that Samsung will demonstrate Bixby integrated with Perplexity for the first time at the S26 series launch conference. The specific implementation is: basic tasks are still handled by Bixby, and complex reasoning and questions that require more "thinking" are responded to by Perplexity. The form is similar to Apple's combination of Apple Intelligence and ChatGPT.

Although Apple relies on external partners for some advanced AI applications, it is reported that it has also invested heavily in self-developed large language models (LLM), with the goal of launching a complex cloud-based reasoning model in 2026. By then, Siri is expected to achieve more complex cross-application task processing, and the overall experience will be closer to Claude and ChatGPT. However, Apple does not plan to launch a standalone chatbot application.