Cloudflare introduces Code Mode for Workers AI, a technical approach to fit entire API schemas into 1,000 tokens, enabling LLM agents to execute precise tool calls with minimal latency.

Cloudflare's platform is a leading provider of internet security and performance solutions, offering insights into web security, content delivery, and DNS management. Through documentation, blog posts, and webinars, Cloudflare provides insights into protecting websites and applications from cyber threats and improving performance. Developers and IT professionals can learn about CDN (Content Delivery Network), DDoS mitigation, and firewall configurations to secure and accelerate web traffic.

Cloudflare

Cloudflare introduces Code Mode for its MCP server, a technique that compresses access to the entire Cloudflare API (2,500+ endpoints) into just ~1,000 tokens — a 99.9% reduction compared to a traditional MCP server that would require 1.17 million tokens. Instead of exposing thousands of individual tools, the server exposes only two: search() and execute(). Agents write JavaScript to query the OpenAPI spec and make authenticated API calls, with code running inside a sandboxed V8 isolate (Dynamic Worker). The approach enables progressive capability discovery, fixed token cost regardless of API size, and safe execution without requiring agent-side modifications. Cloudflare is also open-sourcing a Code Mode SDK so developers can apply the same pattern to their own MCP servers.

Code Mode: give agents an entire API in 1,000 tokens

Comparing approaches to context reduction