OpenCode

Use every RelayRouter model as a custom provider inside the OpenCode coding agent.

OpenCode supports custom [OI]-compatible providers through the AI SDK. Add RelayRouter to your config, export a key, and pick a model.

1. Create a key

Open the API keys page, create a key, and copy the secret. Then export it in the shell where you run OpenCode:

export RELAY_API_KEY="sk-or-v1-..."

2. Add the provider

Add this to opencode.json in your project, or to ~/.config/opencode/opencode.json for every project:

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "relayrouter": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "RelayRouter",
      "options": {
        "baseURL": "https://api.relayrouter.org/v1",
        "apiKey": "{env:RELAY_API_KEY}"
      },
      "models": {
        "deepseek-v4.1-flash": {
          "name": "DeepSeek V4.1 Flash"
        },
        "gpt-6-sol": {
          "name": "GPT-6 Sol"
        },
        "claude-sonnet-4-6": {
          "name": "Claude Sonnet 4.6"
        }
      }
    }
  },
  "model": "relayrouter/deepseek-v4.1-flash"
}

Use exact catalog ids

Each key under models must match an id from GET /v1/models. The display name can be anything. Add only the models you want in the picker.

3. Run it

1
Start OpenCode
Run opencode in your project directory.
2
Pick the provider
Run /models and choose a relayrouter/… model.
3
Send a prompt
Tool calls and streaming use the standard chat completions endpoint.

If OpenCode cannot see the models, confirm the base URL ends in /v1 and that RELAY_API_KEY is set in the same shell. To use a different default, change the top-level model value.

PreviousIntegrations NextSillyTavern