Cloudflare announced an expansion of its Agent Cloud platform, introducing new tools and infrastructure to support the development and large-scale deployment of autonomous AI agents.

The update includes features such as Dynamic Workers for efficient execution of AI-generated code, Git-compatible “Artifacts” storage for managing agent-generated data, and general availability of Sandboxes—isolated environments where agents can run full operating systems. The platform also introduces “Think,” a framework enabling long-running, multi-step agent tasks.

Cloudflare aims to address key challenges in scaling AI agents, including cost, security, and infrastructure limitations, by offering a unified platform that integrates multiple AI models and simplifies switching between providers.

The company said the enhancements are designed to move AI agents from experimental use cases to production-grade applications, supporting a future where millions of autonomous agents operate simultaneously across enterprise environments.
Business Wire