FAQ

Is XNova OpenAI compatible?
Yes, and Anthropic compatible too. Most clients use the OpenAI-compatible base URL:
https://api.xnova.online/v1Claude Code uses the Anthropic format with https://api.xnova.online (no /v1). See Claude Code.
Does my balance work for every group?
Yes. Each API key is bound to one group and billed at official price × group multiplier, all from one balance. To use another route, create another key with a different group.
What is the difference between the two nodes?
They share the same account, keys, and balance. Use api.xnova.online by default and switch to cf-api.xnova.online if the main node is slow.
Why does my client return 401?
Usually the API key was copied incorrectly, has extra spaces, or was deleted or disabled. Create a new key and try again.
Why does it say the model does not exist?
The key's group does not include that model, or the name is misspelled. Check the model list.
Why do requests fail occasionally?
- Browser CORS errors: check
Origin, CORS, and domain settings. - Insufficient balance: top up and retry.
- Upstream errors: retry later, or ask an admin to check the account pool.
- Content policy errors: rephrase the prompt.
Why is there no compression setting for PNG images?
output_compression only applies to JPEG and WebP. It is not sent for PNG.
Are images charged with the text multiplier?
No. Images are billed by size tier; text models use the group multiplier.
