Cloudflare Acquires Deno to Merge celld Into workerd and Simplify Self Hosting Workers

Cloudflare announced on October 9, 2026 that it is acquiring Deno, the JavaScript and TypeScript runtime startup co-founded by Node.js creator Ryan Dahl, and that the entire Deno team will join the company. Financial terms were not disclosed. The move brings one of the best known independent runtime projects of the last several years inside a company that already runs one of the largest edge networks in the world. It also ends, in its current form, a multi-year experiment in building a server side JavaScript runtime from scratch.

Deno was started by Dahl and co-founder Bert Belder after Dahl created Node.js and then grew frustrated with parts of its design. The startup raised $26 million in total, including a $21 million Series A led by Sequoia in 2022, according to TechCrunch. Its runtime, the Deno Deploy hosting service and the JSR package registry became familiar names to JavaScript developers, even if they never displaced Node.js in production.

The acquisition is not really about the Deno runtime itself. The specific piece that caught Cloudflare's attention is celld, an open source implementation of the Cloudflare Workers programming model that the Deno team launched in August. celld packages the Workers model into one binary written in Rust, with object storage as its only external service dependency, which means the same code and APIs that normally run only on Cloudflare's network can be run on a developer's own servers.

Cloudflare's equivalent piece of infrastructure is workerd, the open source runtime that powers Cloudflare Workers and Durable Objects. Durable Objects are distributed singletons that pair single threaded JavaScript execution with their own SQLite database and WebSocket support. The plan, as both companies described it, is to merge workerd and celld so that Workers and Durable Objects can be self-hosted in a far simpler way.

Key Facts

TechCrunch reported on October 10 that Cloudflare announced the acquisition on Friday, October 9, and that the company will use the deal to improve its Workers programming model and platform, which lets customers build and run software on Cloudflare's network. The same report put Deno's total fundraising at $26 million, including the $21 million Series A led by Sequoia that was announced in 2022, and confirmed that the financial terms of the acquisition were not disclosed.

Deno wrote on October 9 that the entire team is joining Cloudflare and laid out concrete consequences for existing users. The Deno runtime will be supported for another year with monthly releases containing bug-fixes and security updates, after which development of the runtime ends. Deno Deploy will continue operating for six months before shutting down, with migration support offered to paying customers moving to Cloudflare Workers. JSR will keep operating, with its infrastructure moving to Cloudflare, and the team will keep supporting rusty_v8 while working to integrate it into workerd.

Cloudflare said in its own post on October 9 that the goal is to radically simplify self-hosting Workers and Durable Objects so developers can use the same primitives in more places. Dahl, now a Senior Principal Engineer at Cloudflare, wrote that the Deno team is joining Cloudflare and that the two runtimes are being merged. Kenton Varda, a Distinguished Engineer at Cloudflare and principal engineer for Cloudflare Workers, added that when the Deno team released celld in August, the Cloudflare team was delighted.

Startup Fortune reported on October 11 that the real prize of the deal is celld, the Rust tool the Deno team shipped quietly in August, and that Cloudflare shares were little changed in the immediate aftermath of the news. The report, citing a Yahoo Finance report, repeated that terms were not disclosed and that Deno had raised $26 million in total.

On the strategic side, Varda pushed back on what he called a popular theory that Cloudflare made Workers work differently from other cloud platforms in order to create lock-in. Cloudflare believes, he wrote, that being open source and giving people an escape hatch is good business.

Analysis

Cloudflare is not buying a runtime that failed to beat Node.js. It is buying a small, focused team that has already written the hardest part of a portable version of its own platform. Startup Fortune reported on October 11 that Dahl concluded Deno never solved the problem it set out to fix, describing the runtime as trapped chasing compatibility with Node.js rather than replacing it. Read that way, the acquisition is less a rescue than a reallocation of a very specific kind of engineering talent toward a problem Cloudflare itself has struggled to solve, namely how to let customers run Workers away from Cloudflare's network.

What this really means is that the center of gravity in serverless computing is shifting from where code runs to what programming model developers write against. Cloudflare has argued for years that its model, built around isolate based execution, Durable Objects and a high level JavaScript interface, is the right abstraction. The missing piece was portability, since a model that only exists on one vendor's edge is easy to dismiss as lock-in. Merging celld into workerd directly attacks that criticism, and it does so with a tool that Cloudflare did not have to build from scratch.

The timing of the Deno wind down is also revealing. A one-year support window for the runtime and a six-month window for Deno Deploy are short by open source standards, but they are consistent with a team that wants to move quickly. JSR surviving and moving to Cloudflare, along with continued work on rusty_v8, suggests Cloudflare wants to keep the parts of Deno's stack that are useful infrastructure while retiring the parts that compete with its own products.

The strategic logic extends beyond runtimes. Both companies framed Durable Objects as a strong fit for AI agent workloads, pointing to cheap serverless execution, persistent state, WebSockets and a high level JavaScript interface. If agents become a major category of software, the vendor that offers both a hosted version and a self-hosted escape hatch may be better positioned than one that offers only a managed service.

Why It Matters

For developers, the practical effect is a set of deadlines. Anyone running the Deno runtime in production has roughly one year of monthly bug-fix and security releases before official development stops, though Deno will remain open source and Cloudflare says it welcomes others who want to continue its development. Teams on Deno Deploy have six months to move, with migration support to Cloudflare Workers. That is a short runway for organizations that built on either product.

For the wider cloud market, the deal tests whether open source self-hosting can be a credible answer to lock-in. Varda's argument that giving people an escape hatch is good business is a direct challenge to the assumption that cloud platforms should keep their runtimes closed. If workerd and celld merge cleanly, customers will be able to run the same Workers primitives on their own infrastructure, which changes how much leverage a provider has over them.

There is also a consolidation story here. Startup Fortune reported on October 11 that Cloudflare has already acquired VoidZero, the steward of the Vite build tool and the Rolldown bundler, as well as Astro. Adding Deno to that list suggests Cloudflare is assembling a broad slice of the JavaScript and TypeScript toolchain rather than making a single point bet.

Next Up

The immediate task is technical. Dahl and Belder are expected to lead the effort to merge celld into workerd, with the stated aim of making self-hosted Workers and Durable Objects simple enough to be practical. Continued work on rusty_v8 and its integration into workerd is part of that path, and the celld project, which shipped in August, is the starting point.

After that, the measure of success will be adoption. Cloudflare wants the Workers programming model to become a mainstream way to build servers, running either on its own network or on a customer's infrastructure. Whether the merged runtime wins over developers who have spent years on Node.js, or on Deno for that matter, will not be clear until the first self-hosted deployments appear.