Connect Chatbox to Nbility
Chatbox can use Nbility through its OpenAI API Compatible mode. Distinguish API Host from API Path: the host is the origin, while the default path already contains /v1.
Configure the provider
Install the app from Chatbox, open Settings → Model, and select OpenAI API Compatible. If your version supports named providers, add one called Nbility.
| Field | Value |
|---|---|
| API key | A key from Token management |
| API Host | https://api.nbility.ai |
| API Path, if present | Keep /v1/chat/completions |
| Model | The exact ID from Models |
If your version has only a Base URL field, read its hint or inspect the final request address to avoid /v1/v1/chat/completions. Do not use the Nbility website origin as the API host.
Select a model and test
Save the configuration and select this provider and model in a new conversation. Send a short greeting and confirm the request in Usage logs.
If the model picker does not contain your model, enter its exact ID as a custom model. Display names, vendor names, and pricing groups are not model IDs.
Troubleshoot
- 404: check whether both Host and Path include
/v1. - 401: update the API key and remove extra whitespace or line breaks.
- The same key works in an SDK: check that the conversation uses the new provider and saved address.
- Images or files fail: test plain text first, then confirm the model supports that input type.
See Connection basics for the complete diagnosis order.
References
Informed by the aicentos Chatbox tutorial (CC BY 4.0), rewritten for Nbility with additional URL checks. See Chatbox's provider guide for Host and Path semantics.