Cline
Configure Cline with an OpenAI-compatible gateway
Cline's OpenAI Compatible provider supports custom gateways. Agent tasks need a model with streaming and tool calling. Prepare your key/group using the shared guide.
1. Install and open Cline
Install the Cline extension from the official site, open a test workspace in VS Code, then open the Cline panel and click ⚙️ Settings.
2. Select the provider
| Field | Value |
|---|---|
| API Provider | OpenAI Compatible |
| Base URL | https://api.vnaipro.app/v1 |
| API Key | A key created on your gateway website |
| Model ID | An ID from Check models |
| Use Azure Identity Authentication | Off |
This configuration uses Chat Completions; do not append /chat/completions to the base URL.
Do not select an OAuth-based OpenAI/Codex preset when using your gateway key.
3. Check model options
- Set Context Window and Max Output Tokens to the model's actual limits when required.
- Enable images and other capabilities only if the model/channel supports them.
- If Plan and Act use separate providers/models, configure both or select shared settings.
- Save and start a new task.
Do not copy another model's context limits just because it comes from the same vendor. A Responses-only model needs a suitable Responses adapter; this OpenAI Compatible setup does not translate protocols.
4. Test the agent
Ask Cline to read and summarize a README in a test workspace. Approve the tool through the UI and check usage history on your gateway website. If chat works but tools fail, check model tool calling and common errors.
Sources: OpenAI Compatible, Installing Cline.