Model gateway

How AIConcept routes AI generation

Every AI request on the platform passes through a central model gateway: a registry decides which models each app may use, the cost is quoted in AICoin before anything runs, and every generation is written to the usage ledger.

Cost-aware routing

Provider selection weighs the credit estimate, quality target, and app-level model assignment before dispatch.

Guarded fallbacks

Unavailable or policy-blocked providers are skipped without exposing raw provider errors to the user.

Usage ledger

Every generation writes a usage record with model, cost, and receipt metadata, visible in your payment history.

Operator governance

Admins assign models per app and can disable a provider; apps consume only what the registry allows.

What you see before you spend

Quote first, charge second. The server decides reservation and settlement, never the browser.

1

Estimate shown first

Each app shows the AICoin quote for your request before anything runs.

2

Reserve, then run

The server reserves the quoted amount, calls the provider, then settles the actual cost.

3

Release on failure

If the generation fails, the reservation is released and the amount returns to your wallet.

Why no public model catalog?

The available model set changes with provider pricing and policy, so a static table would go stale and mislead. Instead, each app shows the models available to your account together with the exact AICoin quote right before you confirm a generation. This page previously displayed a static status table; that content was removed because it did not reflect real data.

Want to try it? Open the app suite and run your first generation — the quote appears before any credit is spent.
Open app suite
AI models - AIConcept