OpenAI Launches GPT-5 Pro, Sora 2, and Voice Models at Dev Day

[madmax]

OpenAI unveiled a wave of product updates at its Dev Day 2025, including GPT-5 Pro, its most advanced language model yet, alongside the new Sora 2 video generator and a low-cost voice AI model called gpt-realtime mini.

The updates expand OpenAI’s developer ecosystem, offering new tools for agent creation and ChatGPT-based app building. GPT-5 Pro targets industries demanding high reasoning accuracy, such as finance, healthcare, and law.

Meanwhile, gpt-realtime mini brings low-latency, expressive voice interactions at 70% lower cost than previous models, while Sora 2 enables AI-generated video creation — powering OpenAI’s new Sora app, a TikTok-style platform for sharing short AI clips.

CEO Sam Altman said voice and video models will be key to “how people interact with AI in the future.”

[Nahid10]

Huge lineup this year — GPT-5 Pro and Sora 2 sound next level.

[EmTeam]

Finally! Third-party app support was long overdue. Big step for usability.