Reflections on Turso as a Database

I’ve been leaning on Turso a lot more, and honestly the developer experience has been wonderful. It feels like a database that actually matches the pace I want to move at. No endless setup, no tripping over config or permissions, just something I can spin up and start using right away.

What makes it stand out is the simplicity. I can define tables quickly, apply migrations without ceremony, and because it’s SQLite under the hood, the mental model stays light. But it’s also distributed, which means I don’t have to think about where I’m deploying—my queries hit close to wherever I’m running code. That combination of familiar + global is rare, and it keeps me moving instead of second-guessing infrastructure.

I also appreciate how easy it’s been to integrate with Drizzle and Prisma. It’s the kind of glue that just works, and when I reach for Turso I don’t feel like I’m making a huge architectural bet. It’s lean enough that I can treat it like scaffolding if the project doesn’t pan out, but sturdy enough that I don’t worry about scaling when things get real.

Most importantly, it hasn’t gotten in the way. I can stay focused on building features, shaping flows, and iterating on product logic instead of babysitting the database. For where I’m at—running multiple experiments at once, all needing structure without overhead—that’s exactly what I need.

Turso has become the tool that helps me keep momentum. And momentum, more than anything else, is the currency that decides whether these projects live or die.