Using a code model in an IDE
With OpenGateLLM, you can serve code models. These models can be used in the following IDEs to be used as an assistant:
- VS Code
- Intellij
- Pycharm
- Zed Editor
- Cursor
Visual Studio Code
Continue is a Visual Studio Code plugin to use AI code agents.
After installation, please configure the plugin in .continue/config.yml with the following content:
name: Local Assistant
version: 1.0.0
schema: v1
models:
- name: OpenGateLLM
provider: openai
model: <YOUR_MODEL_NAME>
apiBase: <YOUR_API_URL>
apiKey: <YOUR_API_KEY>
roles:
- chat
context:
- provider: code
- provider: docs-legacy
- provider: diff
- provider: terminal
- provider: problems
- provider: folder
- provider: codebase
Intellij/Pycharm
ProxyAI is a Intellij/Pycharm plugin to use AI code agents.
After installation, please configure the plugin in ~/.config/ProxyAI/config.yml with the following content:
- Open Tool > ProxyAI > Providers
- Select
Custom OpenAI
3.Add a new configuration+ - Change the configuration:
name: OpenGateLLM
API Key: <YOUR_API_KEY>
URL: <YOUR_API_URL>
model: <YOUR_MODEL_NAME>
Zed
Zed is a code editor providing native custom IA agent providers.
- Open the Zed Agent Panel
- Open the Model drop-drown menu and select "Configure"
- Select the provider OpenAI
- Input your API Key:
<YOUR_API_KEY> - Input the custom API URL :
<YOUR_API_URL> - Then in the main Zed Menu, click on "Open Settings"
- In the
"language models"add the provider"openai":
"language models": {
// maybe other models here
"openai": {
"api_url": "<YOUR_API_URL>", // this should be automatically inserted
"api_key": "<YOUR_API_KEY>", // if not saved by the input from the Agent Panel
"role": "chat",
"available_models": [
{
"name": "<YOUR_MODEL_NAME>",
"display_name": "<YOUR_MODEL_NAME>",
"supports_tools": true,
"max_tokens": <YOUR_MAX_TOKENS> // retrieve by looking at the max_context_length from /v1/models endpoint
}
],
"version": "1"
}
}
Cursor
Kilo Code AI Agent is a Cursor plugin to use AI code agents.
After installation, please configure the plugin with the following steps:
- Open Kilo
- Select
Custom OpenAI - Add a new configuration :
API Key: <YOUR_API_KEY>
URL: <YOUR_API_URL>/v1/chat/completions
model: <YOUR_MODEL_NAME>