Farewell for now, Chrome.

When I joined the Chrome team, I walked into a group that believed the web could be fast, simple, safe, and a joy to build on. Almost 14 years later, I am still in awe of what this team does every day.

Today I want to say thank you to my friends and community, to share what we have built together over the years, and let you know that I am moving on to a new chapter. More on that soon.

My last day working on Chrome will be December 1. After that, I am excited to be moving into a new role at Google. I am not going far, and I will keep cheering for Chrome and for the open web.

Before I go, I want to look back on some of the work I am proud to have been a part of over Chrome’s 17 years, much of it in partnership with the broader web community.

We evolved what the web can do

When I joined Chrome, the web was still mostly thought of as simple web apps on desktop. Over the years, we helped it grow into a platform where people expect app like experiences on every device, including mobile.

Mobile pushed us to rethink fundamentals. We invested heavily in making Chrome fast and battery conscious on phones, and in giving developers the tools to build first class experiences there. Service Workers were a major shift: they gave the web a reliable way to work offline, handle network failures gracefully, and do more in the background. On top of that foundation, we worked with the community on PWAs so teams could deliver installable, engaging applications that feel at home on users’ devices while staying linked and open.

Alongside this, Chrome and the broader ecosystem steadily expanded what is possible in a browser: richer graphics and media, more powerful input and output, improved install and integration experiences, and a long list of capabilities that helped close many of the so called app gaps with native. It has been a long journey to show that the web can be a truly competitive platform for ambitious applications, and I am proud we played a part in that story.

We changed how the industry thinks about web performance

Core Web Vitals helped make user centric metrics mainstream. By focusing on what users actually feel, we ended up saving Chrome users over 30,000 years of waiting time, and businesses saw real gains in revenue and engagement by optimizing for it, including through 2025. That took years of careful measurement, guidance, tooling, and outreach coming together.

On top of that foundation, we worked to make navigation nearly instant when it counts. Back forward cache and modern prerendering turned going back, forward, and ahead into experiences that often feel immediate. Under the hood, those are complex systems; what matters is that the web feels fast.

Chrome DevTools became and stayed the primary debugging tool for the web

From the early Web Inspector roots to a full suite of debugging and profiling tools, DevTools grew into the compass many developers reach for first. We evolved from the classic Elements, Sources, and Network panels into deep performance tooling, built in Lighthouse that became an industry standard for lab quality, and most recently, Gemini assistance woven across the tool suite.

The Chrome DevTools Protocol powers DevTools itself and a large ecosystem of tools built on top. That arc has always been the same: listen to developer pain and turn it into practical features.

We made it a habit to meet developers where they are

With some work, Chrome evolved to better collaborate with frameworks, tools, CMSs, and platforms into everyday practice, baking performance and UX wins into the defaults of tools people already love. That is developer experience at its best. It took time for Chrome to fully embrace the role of abstractions, but I am glad that this is now a core part of the strategy.

With AI models becoming a central part of how developers generate code, I remain bullish on our work with DeepMind on improving Gemini’s code generation quality for the web and on what that unlocks for framework and tooling ecosystems.

We helped define benchmarks that are both meaningful and motivating

Efforts like Speedometer with browser vendors like Apple and the shared governance around it gave engineers real targets to optimize for that users could feel. In 2025, Chrome set a new high on Speedometer 3.1 after a focused push across V8 and Blink. Those wins are fun, but the reason they matter is simple: the web feels snappier, and people save millions of hours in aggregate.

We showed that extensibility and safety can coexist

Chrome extensions continued to flourish, showing that browser extensibility can be balanced with privacy and security. We completed the long transition from Manifest V2 to V3, with MV2 now disabled for all users. MV3 modernized the platform with service worker based backgrounds and a ban on remotely hosted code, while new capabilities like Offscreen Documents restored key DOM access patterns in a safer way.

We also raised important ceilings, such as safe dynamic rules for declarativeNetRequest, so developers could keep shipping powerful, privacy preserving extensions without hacks. I am especially grateful to everyone who carried feedback from the ecosystem back into design changes, implementation fixes, and documentation.

We helped make reliable browser automation mainstream

Headless Chrome unlocked CI and server workflows early on, and the new headless mode aligned headless and headful on the same code path while the old mode was retired. We supported many Headless Chrome clients over the years, the largest possibly being Google Search’s googlebot.

Chrome for Testing made obtaining consistent, hermetic binaries trivial for CI.

On the protocol front, WebDriver BiDi shipping in ChromeDriver has matured steadily, and Puppeteer now supports BiDi in addition to CDP, making cross browser automation far more accessible. The theme has always been the same: lower friction, higher reliability.

We started exploring how AI agents can help developers build for the web

More recently, we shipped Chrome DevTools MCP to help developers tap into the AI coding agent moment. It is still early, but I am excited about how this direction can reduce toil and help more people feel confident building for the web.

We raised the floor for compatibility and clarity

Interop brought engine teams together to focus on the same pain points, while Baseline gave developers a clearer picture of which APIs are safe to use across browsers. That took steady partnership and paid off in fewer surprises for everyone building on the web.

I was especially proud to see teams work with the tooling and IDE ecosystem to bring Baseline directly into the places developers already work.

What I will remember most

The list could go on. But what I will remember most is not a specific feature or launch. It is the way this team and this community build.

You listen to users and developers. You measure. You sweat the details. You make hard calls with empathy. You take the long view, and you still ship with urgency. That culture is why this team keeps moving the web forward.

The next chapter

The next chapter will be even more interesting. AI will change how people browse and how developers build. But the north stars do not change: make experiences fast, keep people safe, reduce friction for developers, and ship the smallest thing that helps and then iterate.

I truly believe there is no place like Chrome. I am grateful I had the chance to contribute a small part to its story, and I am excited to see what the team and the ecosystem do next.

Thank you to everyone who trusted us with their work. Thank you for the thoughtful feedback, the bug reports, the feature requests, the conference hallway chats, the careful code reviews, and the late nights. Thank you to the developers who told us that the web felt a little less mysterious because of the tools and documentation you built and used.

Although I am shifting my focus, I will still be rooting for you, and for the open web we care about. I am looking forward to staying in touch and to finding new ways for our paths to cross.

With gratitude,

Addy