Get another OpenAI-compatible API Key
Missing from the preset list? If the provider speaks OpenAI-compatible Chat Completions, use OpenAI-compatible / custom in WriteX.
Three things you need
| Item | Notes |
|---|---|
| Base URL | e.g. https://your-gateway.example/v1 |
| API Key | Token required by the gateway; local tools may allow a placeholder |
| Model id | Exact name from the gateway docs |
Confirm network access (VPN/firewall) before testing.
In WriteX
- Settings → Model / Provider → OpenAI-compatible / custom
- Fill Base URL, API Key, and default model
- Test & save
Local models
For Ollama and similar tools, Base URL is often localhost, e.g. http://127.0.0.1:11434/v1. Use a model id you’ve already pulled locally.