Klai runs on thousands of open source projects, wired together on our own European servers. Every line of code we did not write ourselves belongs to somebody else. On this page we pay tribute to 72 of the foundational ones Klai is built upon. The rest stay invisible underneath.
Missing from this page: our own application code. Look for that at github.com/GetKlai.
Operating systems, runtimes, databases, encryption. The base Klai is built on.
11 projects
Building blocks for hosting, auth, storage, and the API our product runs on.
12 projects
The services that make Klai actually do what it does: chat, retrieval, transcription, embeddings, reasoning.
17 projects
Everything that renders a pixel on your screen when you use Klai sits on top of these.
14 projects
How we watch our systems, and where customer knowledge lives as plain git-tracked files.
9 projects
What we use every day to write, review, test, and secure the code that becomes Klai.
9 projects
Our Python lockfiles pin dozens of packages. Our Node lockfile pins several hundred more. Most are small libraries that do one thing well: parse a date, sign a JWT, measure a duration, generate a UUID. No logos. No About pages. Often a single maintainer on GitHub Sponsors for the price of a coffee a month.
We run on them every second of every request.
The full inventory, every library, every Docker image, every sub-processor, lives on our tools page .
They wrote a database in 1996, a web framework in 2018, a type checker in a weekend, an embedding model in a research lab. Then they released it. And we picked it up, years or decades later, and built something on top.
That is how modern software works. A small number of people in each layer, quietly maintaining things for thousands of others. If one of them got hit by a bus, a meaningful part of the internet would wobble. We try not to forget that.
If you maintain one of these projects: thank you. If we use your work and your project is missing from this list, drop us a line at hello@getklai.com and we’ll add it.