Skip to main content

​
Overview

OpenCode lets you configure custom model providers through its config file. With AllTokens, you can route OpenCode requests through a single gateway and manage API keys, models, and quotas in one place.

​
Quick setup

Set your API key: 
export ALLTOKEN_API_KEY="YOUR_API_KEY"
Create or update ~/.config/opencode/opencode.json: 
{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "alltoken": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "AllToken",
      "options": {
        "baseURL": "https://alltoken.co/v1",
        "apiKey": "{env:ALLTOKEN_API_KEY}"
      },
      "models": {
        "glm-4": {
          "name": "GLM-4"
        }
      }
    }
  },
  "model": "alltoken/glm-4"
}
Then start OpenCode: 
opencode

​
Notes

  • Use https://alltoken.co/v1 as the base URL.
  • You can switch models by updating model to values like alltoken/deepseek-v3.
  • You can also put opencode.json in a project root if you want per-project config.