Nbility exposes OpenAI-, Anthropic Claude-, and Google Gemini-compatible APIs. In most SDKs, set the Base URL to https://api.nbility.ai (use https://api.nbility.ai/v1 with OpenAI SDKs) and authenticate with an Nbility API key.
Make your first request
Create a key in Token management, then run:
export NBILITY_API_KEY="YOUR_API_KEY"
curl https://api.nbility.ai/v1/chat/completions \
-H "Authorization: Bearer $NBILITY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.4",
"messages": [{"role": "user", "content": "Introduce yourself in one sentence."}]
}'
If that model is unavailable to your token, choose an ID returned by GET /v1/models.
API map
| Category | Documentation | Endpoint |
|---|---|---|
| Basics | Authentication · Models · Errors | /v1/models |
| Conversation | Chat Completions · Responses · Realtime | /v1/chat/completions · /v1/responses · /v1/realtime |
| Data | Embeddings · Rerank | /v1/embeddings · /v1/rerank |
| Audio | Audio | /v1/audio/transcriptions · /v1/audio/translations · /v1/audio/speech |
| Images | Images · Gemini images | /v1/images/* · /v1beta/models/* |
| Video | Seedance | Asynchronous task APIs |
| Native compatibility | Claude · Gemini | /v1/messages · /v1beta/models/* |
Choose a protocol
- For new projects, the OpenAI SDK and OpenAI-compatible endpoints offer the broadest ecosystem support.
- Existing Claude or Gemini applications can keep their native request format.
- Streaming, tool use, vision, audio, and structured-output support varies by model. Check the model catalog and actual response behavior.
For copy-ready SDK configuration, see SDK quickstart.