MCP Server

Synero + Claude Code

Add Synero as an MCP server and query your AI council directly from Claude Code. Four independent answers Claude reconciles itself, zero context switching.

Setup

Get started in minutes

1

Add to MCP settings

Add Synero to your Claude Code MCP configuration. Get your API key at synero.ai/app/api-keys.

// Add to ~/.claude/settings.json or .mcp.json
{
  "mcpServers": {
    "synero": {
      "command": "npx",
      "args": ["-y", "@syneroai/mcp"],
      "env": {
        "SYNERO_API_KEY": "sk_live_your_key_here"
      }
    }
  }
}
2

Use the tool

Claude Code automatically discovers the query_council tool. Ask it to use Synero for any question where you want multi-model verification.

// Claude Code will now have access to the query_council tool
// Just ask: "Use Synero to verify: Is Rust faster than Go for web servers?"

Full Example

Complete integration

// Example MCP tool call (handled automatically by Claude Code)
{
  "tool": "query_council",
  "arguments": {
    "prompt": "What are the trade-offs between microservices and monoliths?",
    "config": {
      "advisorModels": {
        "architect": "gpt-5.4",
        "philosopher": "claude-opus-4-6",
        "explorer": "gemini-2.5-pro",
        "maverick": "grok-4.3"
      }
    }
  }
}

FAQ

Do I need a Synero account?

Yes. You need a Synero API key, which requires a free account and a subscription ($10/month with credits included). Get your key at synero.ai/app/api-keys.

Does this work with Claude Code in VS Code?

Yes. MCP servers work with Claude Code in both the terminal CLI and VS Code extension. The configuration is the same.

What happens when I ask Claude to use Synero?

Claude Code calls the query_council MCP tool, which sends your question to 4 AI models simultaneously. Each responds independently, and Claude receives all four perspectives plus synthesis guidance to reconcile them itself (roughly half the latency). Pass synthesis: true for a server-synthesized answer instead.

Related

Start using Synero with Claude Code

Four AI models. One synthesized answer. Integrated into your workflow.