How the hub dispatches
One entry point; behind it the hub connects every kind of AI capability — Image live, Text / Video / Audio rolling out.
How it works
Three steps: connect one entry, send one request, get results. The hub handles wiring to multiple AIs underneath.
Connect one entry
One API, one slug. No separate keys or SDKs per provider.
Send one request
Drop your prompt into the hub tower; it picks a channel by modality and health, then dispatches.
Get results
If a channel fails it auto-reroutes. You just collect results — transparent metering.
What the hub connects
Not just forwarding — the hub does routing, failover and billing, all behind one entry point.
One entry point
You only ever talk to the hub tower. Which providers sit behind it, or who gets swapped, is not your concern.
Smart dispatch
Picks channels by modality, health and cost. Congestion or failure reroutes instantly.
Prisms billing
Unified Prisms metering. Free daily quota, top up when needed, fully transparent.