SNAP has published a research model from AI text from AI text on mobile devices to enhance some of the functions of Snapchat in the next few months. The company stated on Tuesday that the model could generate high -resolution images in about 1.4 seconds on the iPhone 16 Pro Max.
The new diffusion model is fully executed on the device and reduces calculation costs compared to models that depend on large servers. SNAP points out that models can create “stunning” visual results by “transferring abundant expressions to large and efficient models.”
SNAP plans to produce technology in the coming months to strengthen the AI snap and AI bitmoji background.
The company states that by implementing this in -house technology, it can provide high -quality AI tools in the community at lower operating costs.
“SNAP has a long history of research in modeling and efficiency,” the company stated on a blog post. “We are inspired by the innovation of the industry, which makes AI tools more efficient and affordable, and are looking forward to continuing to contribute to the rapid pace of innovation, especially for Mobile First Experience. I am.
The company did not share the model details, but noted that it was a continuation of long -term investment in AI and ML technology.
In the past, Snap has tapped AI tools from companies such as Openai and Google to strengthen some of the functions, but debuted a unique in -house AI model that can provide users with unique functions. In particular, almost all high -tech companies, including META, have been doing so for the past few years, making it reasonable to invest in AI and develop their own models.
SNAP will release a 2024 profit report on Tuesday for the fourth quarter.