ToolsThe story, in brief

From model to agent: Equipping the Responses API with a computer environment

OpenAI just shipped agent runtime on the Responses API. No more waiting for agentic frameworks—it's live, stateful, and container-native.

Illustration of a transparent lens revealing connected networks across layers of paper.
Exploring the next frontier of AI research.AI illustration by KeyNews
The KeyNews take

Why it matters

OpenAI is collapsing the gap between model capability and deployed agent infrastructure. By baking agents into the Responses API with file handling, tool orchestration, and hosted compute, they're removing friction from the build-to-production path for founders and enterprises.

The key facts

6 to know
  1. Agent runtime ships via Responses API

  2. Includes shell tool and hosted container support

  3. Enables stateful agent execution with file management

  4. Native tool use and state persistence

  5. Production-ready security and scalability built in

  6. Live as of March 11, 2026

Go to the source

OpenAI Blogopenai.com

Publisher excerpt: How OpenAI built an agent runtime using the Responses API, shell tool, and hosted containers to run secure, scalable agents with files, tools, and state.
Read original report
Back to today's editionMore tools news

Keep reading

Related stories

More from Tools