Join the Lorah
broader beta.
Lorah is a local-first desktop workspace where AI specialists on different models work from one reviewed project memory. Your context stays on your machine, and nothing becomes durable memory without your review.
We are opening the beta in small batches so each cohort gets proper founder support. We read every application ourselves.
What you get
Selected testers receive a 90-day Pro licence — the full product, where the usual trial is 14 days. Complete one short feedback check-in around day 30, positive or negative, and we extend it to 180 days.
There is no credit card and no trial that quietly turns into a charge. If you decide Lorah is not for you, nothing happens.
What we ask
Use Lorah on one real project rather than installing it to look around. A workspace that remembers your work only proves itself on work you actually care about.
Then tell us honestly how it went, including what did not work. The extension depends on completing the check-in, not on saying something positive. A tester who tells us Lorah failed on their project is more useful to us than one who says it was great.
Before you apply
- Lorah runs on Windows and macOS. There is no Linux build yet, and MCP is Windows and macOS only.
- You bring your own model access: a provider API key, or a local or private endpoint. You pay the provider directly at their published rates. We do not mark up inference and we take no commission from model providers.
- This is a beta. There are defects and workflows that are not finished.
How we choose
We are admitting around 50 testers in this batch. We prioritise applicants who name a real project, describe a problem that keeps recurring, already work across several AI tools or need local control, and can configure their own model access.
Being enthusiastic is not the qualification. A skeptical person with a real problem is worth more to this programme than an enthusiastic collector of beta accounts.
If you are not in the current batch we will tell you, rather than leave you without an answer.
What we do with your answers
We use the information in the application to evaluate beta fit, issue access, provide support, and learn from the programme. We do not sell it and we do not share it with advertisers. The application takes about a minute.