Cloudflare, the number one provider of connectivity clouds, made an important announcement about expanding its Agent Cloud through the introduction of an infrastructure, security, and developer toolbox for AI agent development and deployment. This move is expected to take AI agents from local experiment prototypes and make them ready for large-scale deployment on Cloudflare’s global platform.
As the AI industry pivots from simple chatbots toward autonomous agents that reason and execute multi-step tasks, traditional infrastructure has become a bottleneck. Existing models-reliant on costly, “always-on” virtual servers-lack the scalability and security required for a future where millions of personal agents operate simultaneously. Cloudflare is addressing these barriers by providing the compute, storage, and sandboxing environments necessary for the next generation of software development.
“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. We’ve spent nine years building the foundation for this with Cloudflare Workers. Today, we are making Cloudflare the definitive platform for the agentic web.”
Also Read: KORE Partners with Kigen to Launch SGP.32-Compliant IoT Connectivity Portfolio
“Cloud agents are quickly becoming a foundational building block for how work gets done, and with Cloudflare, we’re making it dramatically easier for developers to deploy, production-ready agents powered by GPT-5.4 and Codex to run real enterprise workloads at scale,” said Rohan Varma, Product, Codex, at OpenAI.
A Comprehensive Toolkit for AI Agent Development
Cloudflare’s Agent Cloud equips developers with the essential primitives required to build and scale autonomous tools:
Dynamic Workers: High-Efficiency Compute To make agents economically viable, Cloudflare is introducing Dynamic Workers, an isolate-based runtime. Unlike expensive traditional containers, Dynamic Workers spin up in milliseconds to execute AI-generated code snippets or API calls. This model provides secure isolation at 100x the speed of containers, allowing millions of concurrent executions at a fraction of the cost.
Artifacts: Git-Compatible Agent Storage As agents generate massive volumes of code, Cloudflare is launching Artifacts, a storage primitive built for an agents-first era. Git-compatible and capable of supporting tens of millions of repositories, Artifacts gives agents a permanent, accessible home for code and data that remains compatible with any standard Git client.
Persistent Sandboxes: Full OS Capabilities Now in General Availability, Sandboxes provide agents with a dedicated Linux environment. These isolated spaces include a shell and filesystem, allowing agents to install packages, run complex builds, and iterate with the same feedback loops utilized by human developers.
Think Framework: Enabling Long-Lived Tasks To overcome the limitations of short-lived prompts, Cloudflare introduced Think, a framework within the Agents SDK. Think enables developers to build agents that persist through long-running, multi-platform tasks, maintaining context and state over time.
By removing the friction of infrastructure management, Cloudflare is enabling a world where every developer can deploy secure, scalable, and sophisticated AI agents to handle the world’s most complex digital workloads.