Rool is your private AI workspace. Bring your files, ideas and AI together in one place, on your own machine in the cloud. Unlike mainstream AI, Rool never sells your data, shows no ads and never uses your content to train AI. Your data is stored in Finland, under EU law. It's AI that runs on your terms, with full control and privacy. Built for privacy-conscious individuals, teams and creators. Try Rool free and own your AI.





Free Options
Launch Team / Built With

AWS Startup AdvisorPersonalized AWS guidance that meets you where you’re already building
Promoted



Hi, if I understand this correctly, it comes with GDPR-compliance out of the box?
What stands out about Rool IMHO is the ambition beyond chat: a persistent machine that keeps your files, runs software and carries context across tasks. Combining that with EU hosting and self-hosted AI models makes for a compelling alternative to many AI assistants. Congrats on the launch! 🚀
@wulff Thank you, Morten, for supporting us here and in so many other ways. We really appreciate it!
It's been great so far using Rool! Genuinely recommend it
@christophe_trybulowski Thank you! Happy users like you is what motivates us to keep hacking away at Rool.
Rool has been pretty great so far! Keep up the good work, genuinely recommend it.
@boris_salapa Thanks so much, Boris! Really glad Rool has been working well for you, and thank you for recommending it. If there's ever something you'd love to see in Rool, just let us know.
@viktor_hecht Thank you for sharing! I find that a lot of people use Rool for something very similar. Classic car maitenance, bikes, boats, houses. People who make projects out of passions. Great having you as a user.
"a full machine in the cloud" is the part I keep rereading - does that mean every user gets their own dedicated instance rather than a shared multi-tenant backend? that would explain the privacy story a lot more than a policy promise would, but it's also a much more expensive thing to run per user than typical SaaS AI. curious how the pricing holds up once you're actually provisioning real compute per person instead of just metering API calls
@galdayan Yes, the hypervisor gives the hardware-backend isolation boundary between several user machines on one server, no docker containers. Being very clever and efficient here is what lets us get away with offering this for every user.
@jaankasak makes sense, hypervisor isolation is a lot cheaper per-user than people assume once you're not paying the container orchestration tax on top of it. one thing I'd be curious about as a user is the idle cost side - if my machine just sits there between sessions, am I paying for reserved compute the whole time or does it scale down when nobody's using it