July 8, 2026·5 min read

Cloudflare Bought Astro. Here's What Actually Changes.

TL;DR

Cloudflare acquired The Astro Technology Company in January 2026. The whole team joined Cloudflare, and Astro stays open source and platform-agnostic by pledge.
Astro 7 followed in June 2026: compiler rewritten in Rust, Vite 8 with the Rolldown bundler, 15 to 61 percent faster builds, and a new Advanced Routing entrypoint.
Better Auth wasn't part of the deal. It's an independent, VC-backed auth library that happens to have first-class Cloudflare D1 support, which is why it keeps showing up in the same stack as Astro.
The real risk isn't Astro disappearing. It's the same framework and the same host now sharing one owner, and portability pledges from an acquired team have broken before.

Astro runs Unilever’s marketing pages, Visa’s developer docs, and NBC News. None of that stopped a whole engineering team from becoming Cloudflare employees on January 16, 2026, with no purchase price disclosed and no shareholder vote anyone outside the two companies got to see.

That’s the part worth sitting with before the roadmap talk starts. A framework used by hundreds of thousands of developers just changed hands, quietly, the way most infrastructure decisions that matter actually happen.

What did Cloudflare actually acquire?

Cloudflare bought The Astro Technology Company, the entity behind the open-source Astro framework, and brought its entire team on as full-time Cloudflare employees. No financial terms were disclosed. Astro itself doesn’t change license or ownership structure as a project; it’s still MIT-licensed and governed the way it was before, just funded by a different paycheck now.

Cloudflare also committed money to an Astro Ecosystem Fund alongside Webflow, Netlify, Wix, and Sentry, several of which are direct hosting competitors to Cloudflare Pages. That detail matters more than it looks: it’s a signal, not a guarantee, that Astro is meant to stay useful outside Cloudflare’s own platform.

Why does a CDN company want to own a web framework?

Vercel owns Next.js and has spent years turning that ownership into a hosting moat: Next.js features quietly work best, or only work, on Vercel’s infrastructure. Cloudflare buying Astro reads as the same move from the other direction, a hosting company acquiring the framework layer instead of building one from scratch.

Astro’s pitch was always speed through minimal client-side JavaScript, which lines up cleanly with Cloudflare’s own pitch of speed through edge compute. Owning the framework lets Cloudflare ship platform-specific optimizations, like the CDN cache provider that landed in Astro 7, faster than any partnership agreement would allow.

What actually changed in Astro 7?

Astro 7 shipped in June 2026 with its .astro compiler rewritten in Rust, plus a Rust-powered Markdown and MDX pipeline and a new queue-based rendering engine. It also moved to Vite 8, which brings in Rolldown, a Rust bundler replacing both esbuild and Rollup with a single tool. Reported build times are 15 to 61 percent faster depending on the project.

Two additions matter beyond raw speed. Advanced Routing adds a src/fetch.ts entrypoint that gives full control over the request pipeline, and route caching is now stable with experimental CDN cache providers for Netlify, Vercel, and Cloudflare specifically. The breaking change to plan for: Astro 7 requires Node 22.12 or newer.

Does Better Auth have anything to do with this deal?

No. Better Auth is a separate, independently funded project, built by a self-taught Ethiopian developer and backed by a $5 million seed round from Peak XV and Y Combinator. Cloudflare has no ownership stake in it.

It shows up in the same conversation because Better Auth now treats Cloudflare D1 as a first-class database target: pass a D1 binding directly into betterAuth() and it auto-detects the database, no custom adapter required. Combined with 150,000 weekly downloads and adoption among early-stage AI startups wiring up custom auth for proprietary APIs, Better Auth has become the default auth layer for teams already building on Astro and D1, acquisition or not.

What does this mean if you’re already running Astro on Cloudflare?

In practice, less friction. The stack of Astro plus D1 plus Better Auth was already a common combination for edge-native sites; it just lost a seam. Deployment quirks that used to require a workaround or a third-party adapter are more likely to get fixed upstream now that the same company owns both ends.

The tradeoff sits on the other side of that convenience. Framework and host used to be separate companies with separate incentives, which put a natural check on either one drifting too hard toward vendor lock-in. That check is gone now, replaced by a public pledge instead of a structural one.

Is Astro still safe to bet on?

Developer sentiment on Hacker News split along exactly that line. One commenter, nozzlegear, put it plainly: “I’d been using Gatsby for my static websites for a long time, until it got eaten up by Netlify and then sunset; I switched over to Astro at that point, but now I’m getting a sense of déjà vu.” Gatsby didn’t survive its acquisition. That comparison isn’t unfair, it’s the actual precedent.

Cloudflare’s counter-argument is the Ecosystem Fund and the platform-agnostic pledge, plus a practical point: Cloudflare’s business model depends on developers deploying everywhere, not just on Cloudflare, since usage-based Workers revenue only grows if more sites exist to serve. That’s a different incentive structure than Netlify had with Gatsby, though it’s still an incentive structure Astro’s own team no longer fully controls.

The honest answer is that nobody outside Cloudflare knows yet. What’s true today: Astro 7 is faster, Better Auth’s D1 support makes the surrounding stack easier to wire up, and the framework hasn’t asked anyone to pay Cloudflare directly to use it. Whether that stays true is a bet on stewardship, not on code.