Amazon Web Services has struck an agreement to buy DuckLabs B.V., bringing the company behind DuckDB closer to one of the world’s biggest cloud platforms. The deal is expected to close as soon as next month, and DuckLabs’ team will work with AWS to expand and scale DuckDB’s data services globally.
The agreement comes as DuckDB has built strong momentum in open-source data analysis, while a separate development from Accelerated Understanding points to a new scale for AI-based physics modeling. Together, the two announcements show how much attention is moving toward tools that can work with large amounts of data without adding layers of technical overhead.
Why AWS wants DuckDB
DuckDB is an embeddable, high-performance analytical database designed to make data analysis faster, more portable and more accessible to developers. It is often called the analytics equivalent of SQLite, another database known for fitting directly inside applications.
That design lets developers execute complex SQL-based queries inside applications or client environments without managing server infrastructure. Instead of sending every analysis task to a separate database server, developers can place DuckDB inside the software where the work needs to happen.
DuckLabs said DuckDB’s adoption reached more than three million downloads per day at one point. The project has enjoyed tremendous success and traction since DuckLabs was founded just over five years ago, giving AWS access to a database with an established developer following.
“DuckDB is an incredible open-source project with an amazing community; it is broadly used and very much loved by S3 customers today,” Andy Warfield, AWS Vice President and Distinguished Engineer, said.
AWS and DuckLabs began working together in early 2025. Their collaboration included adding support for DuckDB in Amazon S3 Tables and Amazon SageMaker Lakehouse, connecting DuckDB’s embeddable approach with AWS data services.
AWS wants its native analytics capabilities to become faster, simpler and more cost-effective. DuckDB should benefit from AWS’s cloud infrastructure and global reach, while AWS gains a way to bring DuckDB’s data analysis tools to more users.
Open source remains at the center
The acquisition will not move DuckDB’s core engineering team away from its roots. The team will remain based in Amsterdam under co-founders Hannes Mühleisen and Mark Raasveldt. The independent, nonprofit DuckDB Foundation will continue as the project’s overall steward.
AWS will establish a technical advisory board within the DuckDB Foundation and open DuckDB’s extension ecosystem to signed extensions from third-party developers. That change is aimed at giving developers more ways to build on DuckDB while adding a layer for trusted extensions.
“We are committed to the project’s continued development in open source and under the stewardship of the DuckDB Foundation,” Warfield said.
Before the deal, DuckLabs relied on support contracts and feature development partnerships to monetize open-source users. Going forward, AWS will provide funds to expand investments in the open-source ecosystem, giving the project another source of support while preserving its open-source direction.
Mühleisen described AWS’s role as a way to extend DuckDB’s reach without requiring every user to run a database themselves.
“AWS brings the scale and reach to introduce this technology to many more users, including people who may eventually rely on the Duck Stack without needing to operate a database themselves,” Mühleisen said.
A separate push toward enormous AI data workloads
The same developments include the unveiling of a new AI model from Accelerated Understanding. The model can handle 5 trillion pieces of data in a single prompt, a scale described as 5 million times the size of what Anthropic and Google’s flagship models can typically consume.
Accelerated Understanding is presenting the model for physics modeling. The announcement adds a different angle to the AWS-DuckLabs deal: one focuses on making analytics easier to embed inside applications, while the other focuses on handling vast amounts of information for scientific work.
These are separate events, but they share a basic idea. Developers and researchers need tools that can work with large datasets without forcing them to manage every part of the underlying system. DuckDB addresses that need through an embeddable database, while Accelerated Understanding’s model addresses it through a much larger prompt capacity.
For AWS, DuckLabs offers a path to expand an open-source project that already has strong adoption and a close relationship with S3 customers. For DuckDB, the agreement brings cloud infrastructure, global reach and funding for open-source work while keeping its core team and nonprofit stewardship in place.
Based on
