Openai unveiled new API updates in Dev Day on Monday, where GPT-5 Pro has introduced, the latest linguistic style, its new model for Sora 2 video generation, and a smaller and cheapest audio model.
The updates were part of a series of Ads Directed towards Wooing developers to Openai’s environmental system, including launch Agent building tool and The ability to build applications in Chatgpt.
The addition of GPT-5 Pro may appeal to developers to build applications in the field of financing, legal and health care-which needs “high accuracy and depth of thinking”, according to the CEO of Openai Sam Altman.
Altman also noticed that the vocal capabilities will be necessary in the future because it quickly becomes one of the basic methods that people use to interact with artificial intelligence. To this end, Openai launches “GPT-Haltime Mini”, which is a smaller and cheaper audio model in API that supports low-foot and speech flow reactions. The new model is 70 % cheaper than Openai’s Previous advanced audio form But it is considered “the same quality of sound and expression.”
Finally, creators participating in the developer’s ecosystem in Openai can benefit from Sora 2 in a preview of the application programming interface. Openai Sora 2 released the latest audio generator and videos, last week, along with the Sora application, Tiktok competitor Fillped with short videos created from artificial intelligence. The Sora app allows users to create videos for themselves, friends or anything depends on a router, and sharing it through a tiktok style.
“(Developers) can now access the same model that runs the amazing Sora 2 video outputs in your own application,” Altman said.
Sora 2 depends on its previous generation with more realistic and physically consistent scenes with a simultaneous voice and greater creative control – from a detailed camera direction to visual images.
TECHRUNCH event
San Francisco
|
27-29 October, 2025
“For example, you can take the iPhone and Prade Sora to be expanded to a wide, cinematic shot,” said altman. “But one of the most exciting things we were working on is the quality of these new model husbands with visual images, not just speech, but rich speech, surrounding sound, and the simultaneous effects that are based on what you see.”
Sora 2 is placed as a tool for the development of concepts, whether it is a visual starting point for an advertisement based on the general atmosphere of a product, to a Mattel designer that turns a drawing into a game concept – an example in Dev Day which sheds light on Openai deal with a Barbie maker To bring gym to the gaming pipeline.
https://techcrunch.com/wp-content/uploads/2025/10/Screenshot-2025-10-06-at-3.05.51PM.png?resize=1200,527
Source link