OpenAI has introduced the Agents API in public beta, bringing the same harness and infrastructure that powers Codex to developers through a simple, flexible API. The API provides a powerful harness that manages context, uses tools efficiently, and coordinates subagents, with infrastructure that keeps agents running reliably for days in environments where they can work with files, run code, and save intermediate results.

Developers can create a production-ready agent in a single API call by specifying the task, model, tools, and environment. OpenAI hosts and maintains the harness, while users choose from an OpenAI-managed sandbox, their own infrastructure, or sandbox partners. The API offers versioned access to capabilities with each model launch, so developers don't need to rework their harness with every upgrade.

The Agents API partners with ecosystem providers including Blaxel, Cloudflare, Daytona, Digital Ocean, E2B, Modal, Oracle, Runloop, and Vercel, offering fully managed environments or VPC deployments with specific file and secret storage mechanisms and varied CPU, GPU, and memory configurations. OpenAI also introduced the OpenAI hosted sandbox, leveraging the same infrastructure behind Codex and ChatGPT.