Frequently asked questions¶
This section answers the questions most frequently asked about the gateway.
Which budget is used up?¶
Dashboard view
The The usage budget of this workspace is used up. message does not name the scope. The notice banner above the Dashboard view names it:
- ■ Tenant
- ■ Gateway
- ■ Token
- ■ Plan allowance
- ■ Rate limit
A budget renews at the start of the next billing cycle. See Dashboard.
How high is my rate limit?¶
The rate limit applies to each token. The My Account view lists it in the Tokens section in the Rate Limit column as requests for each time window. A gateway carries a rate limit of its own in addition, and the two apply independently of each other. The Rate limit exceeded — please slow down. message does not name which of the two applied. See Managing tokens.
Why are other models available to me than to other users?¶
The gateway through which your request goes decides the available models, not your user account. For a model that is not enabled on the gateway, the gateway answers with The requested model is not available on this gateway. Users who work through a different gateway therefore see a different selection. See Managing gateways.
What does EU model hosting mean?¶
A gateway can require that requests only go to providers that host in the EU. For a model outside this requirement, the gateway answers with This gateway requires EU model hosting. The selected model routes to a provider that is not EU-hosted. The Myra and Mistral providers and the EU region adapters meet the requirement. See Managing gateways.