AI-assisted coding with Continue.dev

This tutorial shows you how to set up Continue.dev, an open-source AI coding assistant, to use evroc Think's hosted models in VS Code. By the end, you'll have AI-powered code completion, chat, and editing capabilities running entirely on European infrastructure.

Why Continue.dev with evroc Think?

Continue.dev is an open-source alternative to tools like GitHub Copilot. It runs directly in VS Code and uses AI models to help you write, debug, and understand code through autocomplete suggestions, chat, and inline editing.

By connecting Continue.dev to evroc Think, you get:

Data sovereignty — Your code and prompts stay on European infrastructure
Open-source models — Use leading models like Kimi-K2, Llama, and Mistral without vendor lock-in
IDE integration — Native VS Code experience with sidebar panel and inline suggestions

Prerequisites

Before you begin, you'll need:

An evroc account
VS Code installed (version 1.70.0 or later)

Create a project

If you don't already have a project to use, create one in the evroc Console.

Open the evroc Console.
Click + Create Project in the left sidebar.
Enter a name for your project (e.g., continue).
Click Create.

You can also use an existing project if you have one.

Create an API key

To connect Continue.dev to evroc Think, you'll need an API key for the Shared Models service.

In the evroc Console, select your project from the left sidebar.
Navigate to Think > Models.
Click + New in the top right corner.
Select Model, then choose the Shared tab.
Enter a name for your key, for example your username.
Select any model with the Code tag — the key works with all Shared Models.
Click Create API Key.
Copy the API key and save it somewhere secure.

Important: The API key is only shown once. If you lose it, you'll need to create a new one.

Install Continue.dev

Install the Continue extension in VS Code:

Open VS Code.
Open the Extensions view by pressing Ctrl+Shift+X (Windows/Linux) or Cmd+Shift+X (Mac).
Search for "Continue".
Click Install on the Continue extension.

Alternatively, press Ctrl+P (or Cmd+P on Mac) to open Quick Open and run:

ext install Continue.continue

Configure the evroc provider

Continue.dev needs a configuration file to connect to evroc Think. Create the file at ~/.continue/config.yaml:

name: evroc Think
version: 1.0.0
schema: v1

models:
  - name: Kimi-K2-Thinking
    provider: openai
    model: moonshotai/Kimi-K2-Thinking
    apiBase: https://models.think.evroc.com/v1
    apiKey: ${{ secrets.EVROC_API_KEY }}
    roles:
      - chat
    capabilities:
      - tool_use

  - name: mistralai/Devstral-Small-2-24B-Instruct-2512
    provider: openai
    model: mistralai/Devstral-Small-2-24B-Instruct-2512
    apiBase: https://models.think.evroc.com/v1
    apiKey: ${{ secrets.EVROC_API_KEY }}
    roles:
      - edit
      - apply

The configuration uses provider: openai because evroc Think implements the OpenAI-compatible API. The apiBase setting points to the Shared Models endpoint, and ${{ secrets.EVROC_API_KEY }} references a secret from Continue's .env file.

The roles setting defines what each model is used for:

chat — Answering questions in the sidebar
edit — Making inline code changes
apply — Applying suggested changes to files

The capabilities setting enables additional functionality:

tool_use — Allows the model to call external tools and functions for enhanced capabilities

Set your API key

Create a .env file in the Continue configuration directory to store your API key:

echo 'EVROC_API_KEY=your-api-key-here' >> ~/.continue/.env

Replace your-api-key-here with the API key you copied earlier. Continue automatically loads secrets from this file when using the ${{ secrets.VAR_NAME }} syntax in config.yaml.

Alternative: You can also set the API key directly in config.yaml by replacing ${{ secrets.EVROC_API_KEY }} with your actual key, but using a .env file keeps secrets separate from configuration.

Try it out

You're ready to use Continue.dev with evroc Think. Open a project in VS Code and try these features:

Chat — Click the Continue icon in the sidebar and ask questions about your code.
Edit — Select code, press Ctrl+I (Cmd+I on Mac), and describe the changes you want.
Add to Chat — Select code, right-click, and choose Continue > Add to Chat to ask questions about the selected code.

Try asking the assistant to do something:

"Explain the structure of this codebase"
"Write a unit test for this function"
"Refactor this to use async/await"

Next steps

Explore Continue's features in the Continue documentation
Learn about the available models in Think Models concepts
Try different models for different tasks — use faster models for autocomplete and more capable models for complex reasoning

evroc User Documentation