
Hugging Face has released Gradio 5, a major update to its platform for building machine learning web applications. This version introduces server-side rendering for faster load times and a modernized design for core components. It also supports low-latency streaming, allowing for real-time applications such as video streaming and speech transcription. Gradio 5 maintains its simple API while enhancing security and adding an AI Playground for app generation. These updates aim to make Gradio a robust framework for scalable, production-ready ML applications.
Read originalThe latest vLLM release replaces the deadsnakes PPA with building Python from source to improve performance.
© GitHub ChangelogGitHub will deprecate GPT-5.2 and GPT-5.2-Codex models by June 1, 2026. Users are advised to update their workflows to supported models.