opencode

opencode talks to Inferam over OpenAI Chat Completions.

With the CLI

inferam opencode

Arguments pass through: inferam opencode run "explain this repo", inferam opencode --session <id>. The CLI passes its config in OPENCODE_CONFIG_CONTENT for that run only. It also pins small_model and enabled_providers, so a repository’s own opencode.json can’t route any call, or your key, to another provider.

By hand

Add a provider to your global ~/.config/opencode/opencode.json:

{
  "$schema": "https://opencode.ai/config.json",
  "model": "inferam/DeepSeek-V4.1-Flash-1017",
  "small_model": "inferam/DeepSeek-V4.1-Flash-1017",
  "provider": {
    "inferam": {
      "name": "Inferam",
      "npm": "@ai-sdk/openai-compatible",
      "options": {
        "baseURL": "https://api.inferam.ai/v1",
        "apiKey": "{env:INFERAM_API_KEY}"
      },
      "models": {
        "DeepSeek-V4.1-Flash-1017": {
          "name": "DeepSeek-V4.1-Flash-1017",
          "reasoning": true,
          "tool_call": true,
          "limit": { "context": 524288, "output": 32768 }
        }
      }
    }
  }
}

Then export INFERAM_API_KEY=ifr_... and run opencode.