TypeSafe Docs, Console and Login
Four official surfaces, one waitlist gate. Here is what each subdomain holds and which of them you can read without an invitation.
typesafe.ai
PublicMain site and waitlist
The launch blog, the pitch, and the early-access signup form. Start here if you have no invitation yet.
docs.typesafe.ai
PublicTypeSafe docs
The TypeSafe docs cover the client and the question types. Reachable without an account, though the API it describes is not.
console.typesafe.ai
Account requiredTypeSafe login
The TypeSafe login surface. This is what people want when they search TypeSafe login — there is no separate portal.
evals.typesafe.ai
PublicWorkflow evaluations
Where the 193.6x and 444.6x workflow figures are published, including the company's own caveats about the methodology.
hello@typesafe.ai
PublicContact
The published contact address for enquiries that the waitlist form does not cover.
TypeSafe login: there is no separate portal
A surprising share of the search traffic around this launch is people looking for a TypeSafe login page. The answer is short: the TypeSafe console subdomain is it. There is no enterprise portal, no separate dashboard product, and no second account system — one TypeSafe console, reached after your early-access invitation lands. Anyone searching TypeSafe login is looking for that one URL.
If you have not been invited, TypeSafe login will not be useful to you yet. The path is the waitlist form on the main site, then an email. The company has not published how long the queue is or how invitations are prioritised, and early reports suggest developers are pulled off it in batches rather than all at once.
Worth noting for anyone evaluating on a deadline: the TypeSafe docs are readable without an account, and the TypeSafe docs are where the real evaluation starts. You can judge whether your problem fits the three question types, whether the 255-option ceiling is a problem, and whether the schema model suits your data — all before you are anywhere near the front of the queue.
TypeSafe docs and console, in reading order
-
01 — The launch blog
On the main site, outside the TypeSafe docs. It explains the category, the training method and the pricing argument, and is unusually honest about what it is not proving. Read this first.
-
02 — The evaluations
The evals subdomain carries the workflow comparisons that produced the headline multiples, along with the company's own caveats about methodology. Read the caveats, not just the numbers.
-
03 — The documentation
The TypeSafe docs cover the client and the question types. The TypeSafe documentation is where you find out whether your problem can be expressed in the shapes on offer.
-
04 — The adapter repository
Build against the interface using ordinary LLM APIs while you wait for TypeSafe login. It gives you the ergonomics but none of the performance, so treat it as scaffolding rather than a trial.
-
05 — The console
Once invited, TypeSafe login gets you here. Swap the backing client, run your labelled sample, plot confidence against accuracy, and set thresholds from what you observe.
That sequence is deliberately front-loaded with reading the TypeSafe docs. Because TypeSafe login is gated and the performance claims are unverified, the cheapest way to find out whether this is worth your time is to establish whether your problem even fits the shape — which costs nothing but an afternoon in the TypeSafe documentation and requires no invitation at all.
The one thing not to do is wait passively. The queue is the constraint, and the integration work is independent of it. Teams that reach TypeSafe login with the schema already designed and a labelled evaluation set ready will know within a day whether the thing works for them. Teams that start from scratch at that point will spend a week finding out what they could have learned before.
Common TypeSafe login and docs problems
The most frequent one is simply arriving early. TypeSafe login exists and responds, but an account only materialises once an invitation has been issued, so a sign-in attempt before then goes nowhere. There is no self-serve signup path that bypasses the queue, and nothing in the TypeSafe docs suggests one is planned.
The second is looking for TypeSafe docs on the wrong domain. Because the name collides with an older Scala company, search results for TypeSafe documentation routinely surface Akka and Play material from a decade ago. The TypeSafe docs you want live on the docs subdomain of typesafe.ai and mention Jev, System One or RLCD on the first screen.
The third is expecting the evals subdomain to be part of the TypeSafe docs proper. It is not — it is a separate publication of the workflow comparisons, with its own methodology notes. Reading it alongside the TypeSafe documentation is worthwhile, because the caveats there qualify the numbers that appear everywhere else.