Claude Desktop Inference

Cursor IDE Settings

{
  "openai.baseUrl": "http://127.0.0.1:5050/v1",
  "openai.apiKey": "sk-nuera-pro-default",
  "openai.model": "nuera"
}

Cline VS Code Settings

{
  "apiProvider": "openai",
  "openAiBaseUrl": "http://127.0.0.1:5050/v1",
  "openAiApiKey": "sk-nuera-pro-default",
  "openAiModelId": "nuera"
}

OLKIL VS Code Settings

{
  "provider": {
    "nuera": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "nuera",
      "options": {
        "baseURL": "http://127.0.0.1:5050/v1",
        "apiKey": "sk-nuera-pro-default",
        "timeout": 300000
      },
      "models": {
        "nuera": {
          "id": "nuera", "name": "nuera", "tool_call": true
        }
      }
    }
  }
}

Merge into ~/.olkil/vscode-opencode-home/opencode.json, add the provider id to enabled_providers, run npm install @ai-sdk/openai-compatible --no-save, then Reload Window.

Generic OpenAI-Compatible Client

Base URL:	http://127.0.0.1:5050/v1
API key:	sk-nuera-pro-default
Model:	nuera
Auth:	Bearer

For any tool with a custom OpenAI-compatible provider (Open WebUI, LibreChat, LangChain/LangGraph scripts, …). Set these three fields and pick Bearer auth.

Opencode (CLI) Settings

{
  "$schema": "https://opencode.ai/config.json",
  "model": "nuera/nuera",
  "provider": {
    "nuera": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Nuera",
      "options": {
        "baseURL": "http://127.0.0.1:5050/v1",
        "apiKey": "sk-nuera-pro-default",
        "timeout": 300000
      },
      "models": {
        "nuera": { "name": "Nuera", "id": "nuera", "tool_call": true }
      }
    }
  }
}

Merge into opencode.json (or global ~/.config/opencode/opencode.json), then launch opencode and select the nuera model.

Kilo Code (CLI / VS Code) Settings

{
  "$schema": "https://app.kilo.ai/config.json",
  "model": "nuera/nuera-pro",
  "provider": {
    "nuera": {
      "options": {
        "baseURL": "http://127.0.0.1:5050/v1",
        "apiKey": "sk-nuera-pro-default"
      },
      "models": {
        "nuera-pro": {
          "name": "Nuera Pro",
          "id": "nuera-pro",
          "tool_call": true,
          "limit": { "context": 200000, "output": 32768 }
        }
      }
    }
  }
}

Merge into ~/.config/kilo/kilo.jsonc (or project .kilo/kilo.jsonc); VS Code users can also use Settings → Providers → Custom provider → OpenAI Compatible.

Smoke-Test & Verification

curl -sS http://127.0.0.1:5050/v1/chat/completions \
  -H "Authorization: Bearer sk-nuera-pro-default" \
  -H "Content-Type: application/json" \
  -d '{"model":"nuera","messages":[{"role":"user","content":"ping"}]}'

Paste this in any terminal to prove the gateway answers with your config — replace the key with your real one if pasting on another machine. Or click Test gateway + key to ping it from this page (a tiny chat request is billed to the embedded key).