Use Poolside models in OpenCode

Poolside provides an OpenAI-compatible Chat API for Poolside models. Poolside-hosted inference enables thinking by default. For self-managed inference, thinking behavior depends on the deployed model and serving configuration.

Watch the setup video

Prerequisites

You have OpenCode installed. For installation options, see Install OpenCode.
You have one of the following:
- A Poolside Platform API key. To create one, go to platform.poolside.ai.
- The base URL for your self-managed model endpoint, and its API key if the endpoint requires one.
- An OpenRouter API key with access to Poolside models. To create one, go to OpenRouter API keys.

Steps

Poolside Platform
Self-managed Poolside inference
OpenRouter

In Poolside Platform, open API Keys.
Click New key.
Copy the API key.
In OpenCode, run the /connect command.
/connect
Scroll to Other.
Enter poolside as the provider ID.
poolside
Enter your Poolside Platform API key.

Add an example Poolside provider to your opencode.json file. The models block below is an example. Use a Poolside model available to your API key. To find model IDs for your access method, see List available models.

opencode.json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "poolside": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Poolside",
      "options": {
        "baseURL": "https://inference.poolside.ai/v1"
      },
      "models": {
        "<poolside-model-id>": {
          "name": "Poolside model"
        }
      }
    }
  }
}

Run the /models command.
/models
Select a Poolside model.
Start a new chat and send a codebase question to confirm OpenCode responds with the selected Poolside model.

If you run Poolside model inference on your own infrastructure, use the model’s OpenAI-compatible endpoint at https://<model-hostname>/v1. Use the API key configured for the endpoint, if it requires one.

Configure Poolside as a custom OpenAI-compatible provider and provide the endpoint’s API key through the POOLSIDE_API_KEY environment variable.

export POOLSIDE_API_KEY="<api-key>"

If your endpoint does not require authentication, skip the export and omit the env field from the configuration below.

opencode.json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "poolside": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Poolside",
      "env": ["POOLSIDE_API_KEY"],
      "options": {
        "baseURL": "https://<model-hostname>/v1"
      },
      "models": {
        "<poolside-model-id>": {
          "name": "Poolside model"
        }
      }
    }
  }
}

env: Environment variables OpenCode reads your API key from. Omit this field for an endpoint that does not require authentication.
options.baseURL: Your model’s OpenAI-compatible endpoint.
models: Model IDs served by your endpoint. Use a model available from your model server.

Run the /models command, select a Poolside model, then start a new chat and send a codebase question to confirm OpenCode responds with the selected Poolside model.

In OpenCode, run the /connect command.
/connect
Select OpenRouter.
Enter your OpenRouter API key. To create one, go to OpenRouter API keys.
Select a Poolside model. OpenRouter may offer free and paid Poolside models. To see current availability, see Poolside models on OpenRouter.
Start a new chat and send a codebase question to confirm OpenCode responds with the selected Poolside model.

​Watch the setup video

​Prerequisites

​Steps

​Related resources

Watch the setup video

Prerequisites

Steps

Related resources