Continue

Continue is an open-source AI assistant for VS Code and JetBrains IDEs. Configured against LibertAI, chat, edit, and apply all run on open-weights models, and switching providers later is a one-line config change.

Prerequisites

• The Continue extension installed in your IDE
• A LibertAI API key — see Get an API key

Configure

Continue is configured through config.yaml (~/.continue/config.yaml, or open it via the Continue panel with settings gear → Open configuration file). Add a LibertAI model using the generic openai provider with a custom apiBase:

models:
  - name: LibertAI Qwen 3.5 122B
    provider: openai
    model: qwen3.5-122b-a10b
    apiBase: https://api.libertai.io/v1
    apiKey: YOUR_API_KEY
    roles:
      - chat
      - edit
      - apply

Save the file; Continue reloads the configuration automatically. You can add several entries (for example a second one with model: qwen3.6-35b-a3b) and switch between them in the model dropdown.

Verify it works

Open the Continue chat panel, select the LibertAI model in the model dropdown, and send:

Say "ready" if you can hear me.

If you get a reply, chat is working. Highlight some code and use Edit to confirm the edit role too. On a 401, check the key directly:

curl -i https://api.libertai.io/libertai/auth/check \
  -H "Authorization: Bearer YOUR_API_KEY"

Models

qwen3.5-122b-a10b is the recommended default for coding (262k context, vision, tools). See the full model and pricing table for budget and TEE options.