Cloudflare, Inc., the connectivity cloud company, has introduced new capabilities to help developers build and deploy AI agents faster and more efficiently. The update includes the industry’s first remote Model Context Protocol (MCP) server, general availability of Workflows, and a free tier for Durable Objects.

These offerings aim to reduce the time and complexity involved in creating AI agents. Developers can now build AI systems in minutes instead of months. The tools are designed to support scalability, maintain affordability, and offer global reach.

AI agents are systems that act autonomously, make decisions, and adapt over time. However, building them requires access to reasoning models, execution workflows, and APIs. Cloudflare addresses all three through its developer platform.

According to Matthew Prince, co-founder and CEO of Cloudflare, “Cloudflare is the best place to build and scale AI agents.” He added that the company’s global network and serverless platform provide developers with the speed and flexibility needed to innovate.

With the new MCP server, AI agents can now complete real tasks, like sending emails or updating code, over the Internet. Previously, MCP was limited to local environments. Cloudflare now allows developers to run MCP servers in the cloud with secure connections. Integration with Auth0, Stytch, and WorkOS simplifies authentication and authorization.

Cloudflare has also expanded access to Durable Objects by offering them on its free tier. Durable Objects allow developers to build stateful applications that retain context across interactions. This is essential for agents that need memory or history to function intelligently. Cloudflare’s network ensures low-latency and high scalability.

Additionally, Workflows is now generally available. It enables developers to build multi-step applications that persist over time. AI agents can now perform complex actions like booking a trip or processing payments with reliable execution.

Cloudflare’s serverless infrastructure also helps reduce costs. Unlike traditional providers, Cloudflare automatically scales resources up or down based on usage. This means organizations only pay for what they use, making AI deployment more cost-efficient.

Kate Holterhoff, senior analyst at RedMonk, noted, “Cloudflare offers a developer-friendly ecosystem for creating AI agents that includes a free tier for Durable Objects and serverless options for AI inference.”

Cloudflare has GPUs in more than 190 cities, allowing low-latency AI inference close to users. This global presence supports real-time performance and efficient scalability.

Related post

View all