VS Code Continue

VS Code Continue

Configure Continue, the open-source AI code assistant for VS Code, to use agentgateway as its LLM backend.

Overview

Continue is a VS Code extension that provides AI-powered code completion, chat, and refactoring. It supports custom OpenAI-compatible endpoints via its configuration file.

Before you begin

• agentgateway running and accessible (e.g., http://localhost:3000).
• A configured LLM backend in agentgateway.
• VS Code with Continue extension installed.

Installation

1. Install Continue from the VS Code marketplace.

   • Open VS Code.
   • Go to Extensions (Cmd + Shift + X on macOS, Ctrl + Shift + X on Windows/Linux).
   • Search for “Continue”.
   • Click Install.

2. Continue will create a configuration file at ~/.continue/config.json.

Configuration

Basic setup

Edit ~/.continue/config.json to add your agentgateway endpoint:

{
  "models": [
    {
      "title": "agentgateway",
      "provider": "openai",
      "model": "gpt-4o-mini",
      "apiBase": "http://localhost:3000/v1",
      "apiKey": "anything"
    }
  ],
  "tabAutocompleteModel": {
    "title": "agentgateway Autocomplete",
    "provider": "openai",
    "model": "gpt-4o-mini",
    "apiBase": "http://localhost:3000/v1",
    "apiKey": "anything"
  }
}

Configuration options

Multi-model setup

Configure multiple models for different use cases:

{
  "models": [
    {
      "title": "GPT-4o via Gateway",
      "provider": "openai",
      "model": "gpt-4o-mini",
      "apiBase": "http://localhost:3000/v1",
      "apiKey": "anything"
    },
    {
      "title": "Claude via Gateway",
      "provider": "openai",
      "model": "claude-sonnet-4-20250514",
      "apiBase": "http://localhost:3001/v1",
      "apiKey": "anything"
    }
  ],
  "tabAutocompleteModel": {
    "title": "GPT-4o Autocomplete",
    "provider": "openai",
    "model": "gpt-4o-mini",
    "apiBase": "http://localhost:3000/v1",
    "apiKey": "anything"
  }
}

Switch between models using the Continue sidebar model selector.

Features

Chat

1. Open Continue sidebar (Cmd + M on macOS, Ctrl + M on Windows/Linux).
2. Type your question in the chat input.
3. Continue sends requests to agentgateway.

Code completion

Continue sends code context to your configured tabAutocompleteModel for inline suggestions.

Slash commands

Use Continue’s slash commands with your agentgateway backend:

• /edit - Refactor selected code.
• /comment - Add documentation comments.
• /test - Generate unit tests.
• /fix - Fix code issues.

Example agentgateway configuration

Here’s a gateway configuration optimized for Continue with multiple backends:

# yaml-language-server: $schema=https://agentgateway.dev/schema/config
binds:
- port: 3000
  listeners:
  - routes:
    - policies:
        backendAuth:
          key: $OPENAI_API_KEY
      backends:
      - ai:
          name: openai-chat
          provider:
            openAI:
              model: gpt-4o-mini

- port: 3001
  listeners:
  - routes:
    - policies:
        backendAuth:
          key: $ANTHROPIC_API_KEY
      backends:
      - ai:
          name: anthropic-chat
          provider:
            anthropic:
              model: claude-sonnet-4-20250514

Environment variables

Continue respects OpenAI environment variables. Set these before launching VS Code:

export OPENAI_API_BASE=http://localhost:3000/v1
export OPENAI_API_KEY=anything
code .

Verification

1. Open Continue sidebar in VS Code.
2. Select your agentgateway model from the model dropdown.
3. Send a test message: “Hello, are you working?”.
4. Continue should respond via agentgateway.

Check agentgateway logs to confirm requests are being received.

Troubleshooting

Connection refused

• Verify agentgateway is running: curl http://localhost:3000/v1/models.
• Check firewall settings.
• Ensure apiBase includes /v1 path.

“Model not found” error

• Ensure the model field matches your agentgateway backend configuration.
• Check agentgateway logs for backend errors.

Slow completions

• Check agentgateway latency in observability metrics.
• Consider using a faster model for tabAutocompleteModel.
• Adjust Continue’s requestDelay setting in config.json:

  {
    "requestDelay": 500
  }

API key errors

• If agentgateway has no authentication, use any placeholder: "anything".
• If using backendAuth, ensure gateway has valid provider API keys.
• Check agentgateway logs for auth errors.

Advanced configuration

Custom context providers

Continue can send additional context with requests. Configure context providers in config.json:

{
  "contextProviders": [
    {
      "name": "code",
      "params": {
        "maxSnippets": 10
      }
    },
    {
      "name": "file",
      "params": {}
    }
  ]
}

Temperature and parameters

Adjust model parameters:

{
  "models": [
    {
      "title": "agentgateway",
      "provider": "openai",
      "model": "gpt-4o-mini",
      "apiBase": "http://localhost:3000/v1",
      "apiKey": "anything",
      "temperature": 0.7,
      "topP": 0.9,
      "maxTokens": 2000
    }
  ]
}

Related documentation

• Continue Documentation
• agentgateway LLM Configuration
• Backend Authentication