The biggest trends in Cloudflare Pages right now

Cloudflare Pages has shifted from a static site host to the primary deployment surface for AI-native applications, driven by a developer base that now exceeds 10 million. This growth is fueled by a surge in agentic traffic, with daily requests from AI agents on the network growing by more than 1,700% over the past year. For the first time, more than half of all Internet traffic is non-human, forcing platforms to support continuous interaction between people, APIs, and autonomous agents. The platform’s architecture, built on Workers isolates rather than traditional containers, allows developers to deploy globally without region selection, ensuring sub-5ms cold starts for dynamic functions. This shift means Pages is no longer just about serving HTML files; it is the infrastructure layer where AI agents execute code and process data at the edge.

Global performance and pricing advantages

The most significant technical trend is the measurable advantage Cloudflare Pages holds in global response times, particularly outside North America and Western Europe. With over 340 cities covered across 130 countries, the network delivers sub-50ms Time To First Byte (TTFB) for over 95% of users globally when serving cached static content. In regions like Asia Pacific, Africa, and the Middle East, this density creates a structural performance gap against competitors with fewer points of presence, directly improving Largest Contentful Paint scores for international audiences. You already know that TTFB drives LCP, so this geographic coverage is the single biggest factor for sites with a global user base.

Pricing remains a major differentiator, with unlimited bandwidth included on both the free and Pro tiers, a feature no other major competitor matches. The free tier allows up to 500 builds per month and 100 sites, while the Pro tier costs $5 per month and adds 5,000 builds and 10 million Workers requests. However, this model requires careful budgeting because Pages Functions consume Workers allowances separately from static asset delivery. A site with extensive dynamic behavior will incur usage-based charges that can exceed the base subscription, meaning the "unlimited" label applies only to file delivery, not compute execution. Developers must monitor CPU time and request counts to avoid unexpected bills, as the starting price is a minimum, not a forecast.

AI integration and new tooling

The second major trend is the deep integration of AI tools into the development workflow, exemplified by the open-sourcing of Kitesurf, a browser engine for AI agents running entirely on Cloudflare Workers. Released under the Apache 2.0 license, Kitesurf parses HTML, applies CSS, and executes JavaScript without a browser container, supporting the Chrome DevTools Protocol for clients like Puppeteer. It runs on V8 isolates with near-zero cold starts, allowing developers to inspect the source, run it locally via Wrangler, or deploy their own instance to their Cloudflare account. This tool is still experimental but represents a clear move toward providing lightweight, serverless environments for agent-driven web interactions.

New endpoints in the Browser Run service further extend these capabilities, enabling programmatic capture of screenshots, PDFs, and structured JSON data from webpages. These features allow developers to build AI-powered scraping and analysis tools directly within the Workers runtime, eliminating the need for external HTTP requests or API tokens. The platform also supports fine-tuning open-weights large language models on edge GPUs with sub-50ms cold starts, offering specific models like Llama 3.3 70B FP8 at $0.29 per million tokens. While these tools provide powerful primitives for building AI-native apps, they require a Workers Paid plan, adding a cost layer for teams that previously relied solely on free static hosting. The question remains whether the complexity of managing these isolated AI workloads outweighs the benefit of keeping everything on a single network.