Use Poolside models in Goose

Watch the setup video

You have Goose CLI installed. These steps use Goose CLI to configure provider settings. Goose shares core provider settings between Goose CLI and Goose Desktop. For installation options, see the Goose Quickstart.
You have one of the following:
- A Poolside Platform API key. To create one, go to platform.poolside.ai.
- An API key for your organization’s Poolside deployment.
- An OpenRouter API key with access to Poolside models. To create one, go to OpenRouter API keys.

Goose splits the endpoint into a host and a base path, rather than taking a single base URL.

In your terminal, run goose configure.
goose configure
Select Configure Providers.
Select OpenAI.
When prompted for OPENAI_API_KEY, paste your Poolside API key.
For OPENAI_HOST, enter https://inference.poolside.ai. Do not include /v1. The base path covers it.
For OPENAI_BASE_PATH, keep the default, v1/chat/completions.
Enter <poolside-model-id>, replacing it with an available Poolside model ID. To find model IDs for your access method, see List available models. Goose checks the configuration and saves it. To rename the provider or connect multiple OpenAI-compatible endpoints, create a custom provider instead. See Configure LLM Provider in the Goose documentation.
Start an interactive session from your project directory. In Goose CLI, run:
goose session
Send a codebase question to confirm Goose responds with the selected Poolside model.

Goose splits the endpoint into a host and a base path, rather than taking a single base URL.

In your terminal, run goose configure.
goose configure
Select Configure Providers.
Select OpenAI.
When prompted for OPENAI_API_KEY, paste the API key for your Poolside deployment.
For OPENAI_HOST, enter https://<api-domain>. Do not include /openai/v1. The base path covers it.
For OPENAI_BASE_PATH, enter openai/v1/chat/completions.
Enter <poolside-model-id>, replacing it with an available Poolside model ID. To find model IDs for your access method, see List available models. Goose checks the configuration and saves it. To rename the provider or connect multiple OpenAI-compatible endpoints, create a custom provider instead. See Configure LLM Provider in the Goose documentation.
Start an interactive session from your project directory. In Goose CLI, run:
goose session
Send a codebase question to confirm Goose responds with the selected Poolside model.

Use this option if you already route your models through OpenRouter, or want to compare Poolside models alongside others.

Create an API key in your OpenRouter account settings.
In your terminal, run goose configure.
goose configure
Select Configure Providers.
Select OpenRouter.
When prompted, paste your OpenRouter API key.
Enter <poolside-model-id>, replacing it with an available Poolside model ID. OpenRouter may offer free and paid Poolside models. To see current availability and model IDs, see Poolside models on OpenRouter.
Start an interactive session from your project directory. In Goose CLI, run:
goose session
Send a codebase question to confirm Goose responds with the selected Poolside model.