FAQ
https://17astra.com/v1
Does /v1/models guarantee every model works?
The list shows models configured for the current credentials. Check the exact model ID and its supported endpoint. A successful end-to-end request and a correct usage record are required to verify operation; a listing alone is not a successful test.
Which Base URL should I enter?
Usually https://17astra.com/v1. If your client automatically appends /v1, use https://17astra.com instead. Avoid /v1/v1. Use an API token created in your own Astra-api account.
Why can an account with a balance still be rejected?
The token can have its own budget, expiration, model restrictions or IP restrictions. Check both the account balance and the token settings. A token budget does not add money to the account.
Are audio models tested in text chat?
No. Text-to-speech requires the audio endpoint and compatible input fields. Check the current service status before using audio. A blank text-chat bubble does not establish that an audio model is broken or working.
Can I enable payments myself?
Online payments remain disabled in this launch phase. Credentials, callbacks, reconciliation, duplicate-payment protection and refund handling must be configured and tested before payments can be enabled.