An increasing number of companies are beginning to develop their own generative artificial intelligence models. Notable examples include Google, Meta, and OpenAI, which have created some of the most advanced models in this space, such as Google Gemini, Meta’s Llama, and OpenAI's GPT. However, only a few consumer-facing, hardware-centric brands have ventured into building their own models. Samsung is one of them.
In 2023, the company introduced its generative AI model, Gauss. And now, at its latest developer conference, Samsung unveiled the successor to this model, now called Gauss 2. That said, this model is not yet integrated into Samsung’s Galaxy ecosystem, such as its smartphones like the Galaxy S24 Ultra. For now, Gauss 2 appears to have a different use case altogether, including internal applications to facilitate various tasks and assist employees.
“Samsung Electronics is committed to developing cutting-edge software, including AI and data analytics, to enhance user experiences,” said Paul Kyungwhoon Cheun, President, CTO of the Device eXperience (DX) Division and Head of Samsung Research. “With three distinct models, Samsung Gauss2 is already boosting our internal productivity, and we plan to integrate it into products to deliver higher levels of convenience and personalisation,” he added.
Also Read: iPhone 17 Pro will exclusively get this big camera feature, standard models to miss out: Report
Samsung says that Gauss 2 is both more efficient and more powerful than the first-generation large language model. And like some of the more advanced models, such as Google's Gemini, it is available in multiple flavours: compact, balanced, and supreme. These models are designed to power different kinds of use cases and applications.
As the name suggests, the compact model has been designed to power on-device use cases where computing resources may be limited. Balanced, of course, focuses on achieving a balance between performance, speed, and efficiency. But the Supreme model is the most powerful and capable, and is said to have reduced computational costs during training and inference processes while maintaining high levels of both performance and efficiency.
Samsung says Gauss 2 is a multimodal model and supports 9 to 14 different languages, including English, Korean, and more. “Samsung Gauss2 supports 9 to 14 languages as well as various programming languages, depending on the model. Samsung has developed and incorporated its own stabilisation techniques for training large language models (LLMs) and designed a custom tokenizer to ensure maximum efficiency for these supported languages,” Samsung said.
So far, Samsung has been using the model internally. “Since its launch last December, the monthly usage of code.i has quadrupled, with about 60% of all software developers in the DX Division now using it,” Samsung said.
Additionally, the company noted that it is used in an AI service that helps employees within the DX Division with tasks such as document summarisation, translation, and email composition.
Several reports suggest that Samsung could use this model to power various use cases in its Android 15-based One UI 7 update as well. However, Samsung has not confirmed this.
