Flat pricing is not a discount — it is an architecture. Tsunami runs its own routing layer across a pool of provider accounts, so cost is a function of our infrastructure rather than your keystrokes.
Requests are routed across many provider accounts, with automatic fallback when one is slow, rate-limited or unhealthy. Adding capacity means adding an account to the pool — which is why we can offer flat pricing while everyone else meters tokens.
Capacity is added by adding accounts — not by raising your bill.
Some teams have to use a specific model for policy reasons. That is fine — you keep everything else.
Point Tsunami at your own Anthropic account and keep every room, memory, advisor and agent exactly as it is.
Same deal. The platform around the model is what makes Tsunami useful, and it does not change.
Move between managed and your own without losing a conversation. The price does not change either way.
A shared workspace is worthless if one heavy job blocks everyone else.
Every person gets their own lane. A teammate queueing eight long jobs cannot push you to the back — work is dispatched round-robin, so a quick question stays quick.
Two people saving the same shared room at the same moment both keep their work. Sessions merge per conversation rather than one overwriting the other.
Close the tab or press Esc and the server drops the job immediately instead of grinding on for nobody.
Errors are captured, grouped and tracked automatically — so problems surface to us before a customer has to report one.
Every room, every advisor, every project. No token meter, no per-seat maths.