OpenAI ramps up developer push with more powerful models in its API 

Openly unveiled new API updates on his Dev Day on Monday, the GPT -5 Pro, its latest language model, its new video generation model Sora 2 and a small, cheap voice model.

The updates were part of a series Declaration Prepared toward the developers toward the opeina ecosystem, including an introduction Agent-building equipment And The ability to create apps in ChatzipateThe

Adding GPT -5 Pro can be applied to the developers of the applicable applications in finance, legal and healthcare -“high accuracy and depth of logic” to Sam Altman, CEO of OpenAE.

Ultman also points out that in the future voice capabilities will be essential because it quickly becomes one of the primary ways that people use to contact AI. To that end, the OpenA API is launching a small, cheap voice model “GPT-Realtime Mini” that supports short-latacy streaming interactions for audio and speech. The new model is 70% cheaper than Openai Model in the previous advanced voice However “the same voice promises to express the quality and feelings of the same voice.”

To the end, the creators involved in the OpenAI developed ecosystem can now tap the API to Sora 2. Open Sora 2, its latest audio and video generator Sora 2 published, along with the Sora app last week, A tiktok contestant Short AI-filled videos filled. Sora app allows users to make videos on the basis of their own, friends or prompts and share it through the algorithmical feed of the tiktok-style.

“[Developers] Now the same model has access to Sora 2’s stunning video outputs directly strengthen your own app, “said Ultman.

The previous generation of Sora 2 is made with more realistic, physically consistent scenes with synchronized sound and greater creative control – up to the stylized visual from the camera’s direction.

“For example, you can take the iPhone view and prompt it to Sora to extend it to a broom, cinematic wide shot,” said Ultman. “But one of the most interesting things we are working is that this new model is not just speech with Visuals, not just speech, rich soundscapes, enclosed audio, synchronized effects that you are based on” based on what you see ”

The concept of Sora 2 is made as a tool for development, it is based on the general view of a product, whether it be visual initial point for an advertise OPENDE To bring generator AI to toy making pipeline.