Cloudflare Inc. (NYSE: NET) announced the expansion of its Agent Cloud platform with new infrastructure and developer tools designed to help build and deploy AI agents at scale.

The connectivity cloud company introduced several new features including Dynamic Workers, a runtime environment for executing AI-generated code; Artifacts, a Git-compatible storage system; and Think, a framework for building persistent agents that can handle multi-step tasks.

Dynamic Workers uses an isolate-based runtime that the company states can spin up in milliseconds and run JavaScript code more efficiently than traditional containers. The system is designed to handle millions of concurrent executions for agent tasks like API calls and data transformation.

Artifacts provides Git-compatible storage specifically built for AI agents, allowing developers to create repositories and giving agents access to code and data storage. The company also announced general availability of Sandboxes, which provide persistent Linux environments where agents can perform development tasks.

"The way people build software is fundamentally changing. We are entering a world where agents are the ones writing and executing code," said Matthew Prince, co-founder and CEO of Cloudflare. "But agents need a home that is secure by default, scales to millions instantly, and persists across long-running tasks."

The expansion includes an enhanced model catalog following Cloudflare's acquisition of Replicate, offering access to proprietary models from OpenAI including GPT-5.4, as well as open-source alternatives through a unified platform.

Rohan Varma, Product at OpenAI's Codex division, stated that cloud agents are becoming foundational for enterprise work and that Cloudflare's platform enables deployment of production-ready agents powered by GPT-5.4 and Codex for enterprise workloads.

The announcement comes as the AI industry shifts focus from chatbots toward autonomous coding agents and tools that can perform complex, multi-step actions independently.