Continue
Continue is an open source assistant for VS Code and JetBrains IDEs. Its config.json accepts any OpenAI-compatible provider, so LivRouter becomes a chat, edit, or autocomplete model with one block.
Setup
- Install the Continue extension for your IDE.
- Open the Continue config file:
~/.continue/config.json(the Continue panel has an open-config action). - Add LivRouter as a model provider:
{ "models": [ { "title": "GLM 5.3 via LivRouter", "provider": "openai", "model": "glm-5.3", "apiBase": "https://api.livrouter.com/v1", "apiKey": "your-livrouter-api-key", "roles": ["chat", "edit"] } ]}Set model to any current ID from the model catalog. Remove "edit" from roles if you only want the model in chat, or keep both so edit commands route through the same provider.
Verify the setup
Reload the IDE, pick the model you added in Continue's model picker, and send a prompt. The request shows up in your LivRouter usage logs with the model that served it.
Troubleshooting
- 401 or unauthorized — use a LivRouter key from the API keys page, and make sure
apiBaseends with/v1. - Model not found — re-copy the exact ID from the catalog; the field is case-sensitive.
- Config not picked up — reload the window after saving
config.json; Continue reads the file at startup.
For request-level details, see the API quickstart.