Remote Models

To configure a remote model provider during setup, click Use model server.

To configure a remote model provider after setup, navigate to Sidekick -> Settings -> Inference, then scroll down to the Remote Model section.

Configuring an Endpoint

Sidekick works with OpenAI compatible APIs.

To configure an endpoint, get the endpoint URL from your provider, and enter all components that precedes /v1. For example, if the URL is https://api.openai.com/v1, enter https://api.openai.com/ into the Endpoint field.

Next, enter your API key. This is encrypted with a key securely stored in your keychain.

Selecting Models

You can choose 2 remote models, a main model and a worker model. Specified model names must be the same as that listed in your model provider's API documentation.

Main Model

This is the main model that powers most work in Sidekick, such as chat, most tools and more.

Worker Model

The worker model is used for simple tasks that demand speed and responsiveness, but can accept trade-offs in quality. This includes automatic conversation titles generation and commands in Inline Writing Assistant.

Ideally, a worker model should be fast and cheap to run. As a result, reasoning models are not recommended.