Any OpenAI-compatible app
Most AI apps support an OpenAI-compatible provider. Fill in the three fields below and every NezhaGate chat model is yours.
Fill in these three fields
| Field | What to enter |
|---|---|
| Base URL / API host | https://nezhagate.com/v1 |
| API key | A key created on the API Keys page of the console (it starts with sk-). |
| Model | A model ID, such as gpt-5.5. The full list is in the API reference, or call /v1/models. |
Which URL exactly?
Apps differ in what they expect. Go by the app's hint and its error message:
- The app asks for a base URL, API base or proxy URL: enter
https://nezhagate.com/v1 - The app adds /v1 by itself (the error shows /v1/v1/): enter
https://nezhagate.com - The app wants the full endpoint (API URL, endpoint): enter
https://nezhagate.com/v1/chat/completions
Check the key from a terminal first
If this command returns the model list, the URL and key are fine and the problem is in the app's settings.
curl https://nezhagate.com/v1/models -H "Authorization: Bearer YOUR_API_KEY"
Common errors
- 401: the key is mistyped, has a stray space, or is disabled.
- 404: wrong URL. Try another of the three forms above.
- 400 model_not_found: the model name is misspelled, or the model has been retired.
- 402: not enough balance, or the key has reached its budget.
Images and video
The image features in chat apps usually wait for a synchronous OpenAI-style answer, while NezhaGate's image and video endpoints are async jobs that return a job id first, so those features usually do not work. For images and video, call the API directly or use the Playground on the website.
Apps with step-by-step guides
We tested these step by step; just follow along:
- Cline: Code in VS Code with DeepSeek, Kimi, GLM, Qwen or Claude.
- Cherry Studio: Desktop AI client: add a provider, pick models, start chatting.
- SillyTavern: The role-play chat front end, connected to NezhaGate step by step.
Claude-only tools
For tools that only speak the Anthropic API, such as Claude Code, see Claude Code.
