Skip to main content
Continue uses NagaAI as an OpenAI provider with a custom base URL. For model ids outside OpenAI’s own list, Continue sends Chat Completions and has no setting for Responses. You can assign a NagaAI model to the chat, edit and apply roles. The configuration follows the Continue docs.

​
Before you start

You need a NagaAI API key and the Continue extension in VS Code or JetBrains.

​
Connect Continue

1

Add the model

Add NagaAI as an OpenAI provider in Continue’s config.yaml:
2

Store the key

Put the key in ~/.continue/.env as NAGA_API_KEY=.... The IDE extensions don’t read shell variables.
3

Check the connection

Open the Continue chat, pick the model and send a short message. The request appears in Logs in your dashboard.

​
Capabilities and autocomplete

Agent mode needs capabilities: [tool_use], because Continue doesn’t know the capabilities of model ids outside its own list. NagaAI has no legacy /v1/completions endpoint, so autocomplete setups that use it don’t work.

​
Troubleshooting

​
401 invalid_api_key

Continue found no key. Check that ~/.continue/.env holds NAGA_API_KEY.

​
Agent mode doesn’t call tools

The model entry lacks tool_use under capabilities. Add it to the model entry.