Use OpenCode with the Gateway¶

OpenCode is an open-source terminal coding agent that supports multiple AI providers. Connect it to the Gateway to use Claude and other models through your regional endpoint.

Prerequisites¶

• OpenCode installed: curl -fsSL https://opencode.ai/install | bash or npm i -g opencode-ai
• An LLM API key

Set your API key¶

Add to your shell profile (~/.zshrc, ~/.bashrc, or ~/.config/fish/config.fish):

export AMAZEEAI_API_KEY="your-api-key-here"

Then reload your shell:

source ~/.zshrc   # or ~/.bashrc

Connect OpenCode to the Gateway¶

Create ~/.config/opencode/opencode.json (all projects) or opencode.json in your project root (one project):

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "amazeeai": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Gateway",
      "options": {
        "baseURL": "https://llm.[your-region].amazee.ai/v1",
        "apiKey": "{env:AMAZEEAI_API_KEY}"
      },
      "models": {
        "claude-4-5-sonnet": { "name": "Claude Sonnet 4.5" },
        "claude-4-5-haiku": { "name": "Claude Haiku 4.5" }
      }
    }
  },
  "model": "amazeeai/claude-4-5-sonnet",
  "small_model": "amazeeai/claude-4-5-haiku"
}

Replace [your-region] with your region code (e.g. ch103, us103).

OpenCode only shows the models you list under models. It doesn't read the Gateway's model list. To add a model, copy its exact ID from Available Models or /v1/models. Set small_model too: OpenCode uses it for things like session titles.

Check it's using the Gateway¶

opencode models amazeeai

You should see the models you listed. Then run:

opencode run "say hi"

The header line should show claude-4-5-sonnet. If it shows big-pickle or another opencode/... model, your config wasn't loaded.

Start OpenCode¶

cd /path/to/your/project
opencode

Switching models¶

Pick a model with /models inside OpenCode, or pass -m amazeeai/<model-id>. Only models listed in your models block appear.

Troubleshooting¶

Answers come back but nothing shows in your usage

OpenCode isn't reading your config and fell back to its own models. Check the file is named opencode.json (no leading dot) and is in ~/.config/opencode/ or your project root.

"Model not found" or "Unexpected server error"

The model ID isn't in your models block, or doesn't match the Gateway's ID exactly. Gateway IDs look like claude-4-5-sonnet, not claude-sonnet-4-5.

Auth errors

Run echo $AMAZEEAI_API_KEY in the same shell you start OpenCode from.

"team not allowed to access model"

Your key doesn't include that model. Pick one your /v1/models call returns.