{"id":23116,"date":"2026-08-07T16:29:15","date_gmt":"2026-08-07T16:29:15","guid":{"rendered":"https:\/\/scannn.com\/introducing-kitesurf-the-agent-first-browser-that-runs-in-v8-isolates-on-cloudflare-workers\/"},"modified":"2026-08-07T16:29:15","modified_gmt":"2026-08-07T16:29:15","slug":"introducing-kitesurf-the-agent-first-browser-that-runs-in-v8-isolates-on-cloudflare-workers","status":"publish","type":"post","link":"https:\/\/scannn.com\/lv\/introducing-kitesurf-the-agent-first-browser-that-runs-in-v8-isolates-on-cloudflare-workers\/","title":{"rendered":"Introducing Kitesurf: The agent-first browser that runs in V8 isolates on Cloudflare Workers"},"content":{"rendered":"\n
\n

Should we build our own browser? <\/p>\n

This is one of those questions that has come up every few months internally at Cloudflare for years. Unsurprisingly, it\u2019s the kind that triggers long threads with multiple reasons and persuasive arguments on why we should do it. The browser is obviously the most important software we use every day on our computers; it\u2019s arguably the operating system of the Internet. We\u2019re a company on a mission to help build a better Internet<\/u><\/a> \u2014 who wouldn\u2019t want to take on the challenge of building a new browser?<\/p>\n

But we never quite found the balance between the technical difficulty of such an endeavour and the unique problems we\u2019d be solving by doing it. And so, the idea was shelved, over and over again. Until now.<\/p>\n

Something magical happened: we reached a tipping point where a series of powerful technical advancements in our Developer Platform<\/u><\/a> became a reality, while the advent of AI agents and the demand for a new kind of browser became critical at the same time.<\/p>\n

Running WebAssembly (Wasm)<\/u><\/a> in Workers is now very mature. Primitives like dynamic workers<\/u><\/a>, SQLite-based Durable Objects<\/u><\/a>, Worker-to-worker RPC<\/u><\/a>, service bindings<\/u><\/a>, higher NodeJS compatibility<\/u><\/a> and higher limits<\/u><\/a> open doors to much more ambitious and complex applications that were simply not possible before.<\/p>\n

Browser Run<\/u><\/a>, our headless browser automation API product, has seen tremendous growth with the rise of AI. Agents need browsers in order to perform many tasks, and in many cases cannot succeed without them.<\/p>\n

But there’s a problem \u2014 browser engines like Chromium were built for humans, not agents, and they come with overhead that AI models simply do not need. They consume so much memory and compute that providing every agent with its own instance is prohibitively expensive, restricting large parts of the Web to only the most sophisticated and costly AI models with higher parametric knowledge, while locking out many other agentic applications.\u00a0<\/p>\n

We should be giving all <\/em>agents a browser that excels at what\u2019s important for an AI model, even if that means being light on what\u2019s only useful for humans. For example:<\/p>\n