OpenCode Setup

npm install -g opencode-ai

Or scoop/choco on Windows, brew on macOS, or the install script.

Open OpenCode and run:

/connect

Select Other, enter provider ID:

unlimited-ai

For the API key, enter any value (not used):

sk-any-value

This stores the credential. Now add the provider config below.

Add this to your OpenCode config:

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "unlimited-ai": {
      "name": "vaalory",
      "npm": "@ai-sdk/openai-compatible",
      "options": {
        "baseURL": "https://chat.vaalory.com/api/v1"
      },
      "models": {
        "gateway-gpt-5-5":       { "name": "GPT-5.5" },
        "gateway-gpt-o3":        { "name": "o3" },
        "gateway-gpt-o4-mini":   { "name": "o4-mini" },
        "gateway-gpt-4o":        { "name": "GPT-4o" },
        "gateway-gpt-5-mini":    { "name": "GPT-5 Mini" },
        "gateway-gpt-5-nano":    { "name": "GPT-5 Nano" },
        "gateway-claude-opus-4-7":  { "name": "Claude Opus 4.7" },
        "gateway-claude-sonnet-4":  { "name": "Claude Sonnet 4" },
        "gateway-gemini-3-1-pro":   { "name": "Gemini 3.1 Pro" },
        "gateway-deepseek-v4-pro":  { "name": "DeepSeek V4 Pro" },
        "gateway-deepseek-r1":      { "name": "DeepSeek R1" },
        "gateway-grok-4":           { "name": "Grok 4" },
        "gateway-qwen-3-max":       { "name": "Qwen 3 Max" },
        "gateway-llama-3-3-70b-versatile": { "name": "Llama 3.3 70B" }
      }
    }
  }
}

Config: global ~/.config/opencode/opencode.json or per-project opencode.json

In OpenCode, run:

/models

• Ctrl+M — model picker
• /model — pick model from chat
• Escape — close picker
• Ctrl+N — new chat