MCP Server

ORVEXA MCP Server

Give your AI agents and coding tools real-time access to ORVEXA's model catalog, pricing, and account balance through the Model Context Protocol.

Quick Start

1

Get your API key

Create an API key in the dashboard.

Get your API key
2

Add the server URL

Add this MCP server URL to your client's configuration.

https://api.orvexaproject.com/mcp
3

Start chatting

Start your agent and ask it to list ORVEXA models or check pricing.

All requests use your ORVEXA API key as a Bearer token. MCP usage is billed at the same rates as regular API calls.

Cline (VS Code)

Cline is an autonomous coding agent for VS Code. Add ORVEXA as an MCP server and Cline can call ORVEXA models directly from your editor.

  1. Install the Cline extension in VS Code.
  2. Open the Cline panel and click the MCP servers icon.
  3. Click "Edit Configuration" and paste the JSON below.
  4. Reload the window - the ORVEXA tools will appear.

MCP server configuration

{
  "mcpServers": {
    "orvexa": {
      "type": "http",
      "url": "https://api.orvexaproject.com/mcp",
      "headers": {
        "Authorization": "Bearer ORVEXA_API_KEY"
      }
    }
  }
}

Also: use ORVEXA as Cline's Brain

Want Cline to call ORVEXA's OpenAI-compatible API directly? Create a "Brain" provider with these settings:

  • • Provider: OpenAI Compatible
  • • Base URL: https://api.orvexaproject.com/v1
  • • API Key: your ORVEXA key

Cherry Studio

Cherry Studio supports MCP over Streamable HTTP. Add ORVEXA in the MCP server settings.

  1. Open Cherry Studio and go to Settings, then MCP Servers.
  2. Click "Add Server" and choose the Streamable HTTP transport.
  3. Paste the configuration below and save.

Cherry Studio configuration

{
  "mcpServers": {
    "orvexa": {
      "name": "ORVEXA",
      "type": "streamableHttp",
      "url": "https://api.orvexaproject.com/mcp",
      "headers": {
        "Authorization": "Bearer ORVEXA_API_KEY"
      }
    }
  }
}

Continue

Continue runs in VS Code and JetBrains. Add ORVEXA to your config.json to let your agent call ORVEXA models over MCP.

// Continue config.json (VS Code / JetBrains)
{
  "mcpServers": [
    {
      "name": "orvexa",
      "transport": {
        "type": "streamable-http",
        "url": "https://api.orvexaproject.com/mcp",
        "headers": {
          "Authorization": "Bearer ORVEXA_API_KEY"
        }
      }
    }
  ]
}

Claude Code & Cursor

Prefer the command line or an IDE? Both Claude Code and Cursor can add the ORVEXA MCP server in seconds.

Claude Code (CLI)

# ~/.claude.json (Claude Code CLI)
claude mcp add --transport http orvexa https://api.orvexaproject.com/mcp   --header "Authorization: Bearer ORVEXA_API_KEY"

Cursor

{
  "mcpServers": {
    "orvexa": {
      "url": "https://api.orvexaproject.com/mcp",
      "headers": {
        "Authorization": "Bearer ORVEXA_API_KEY"
      }
    }
  }
}

Tools

The server exposes a set of tools your agent can call automatically.

{model_id}

deepseek-chat - chat with DeepSeek V3 (example) Other active models appear as tools named after their model ID. Each tool takes a required "prompt" plus optional "system", "temperature", "max_tokens" and "stream" arguments.

Resources

Read-only resources your agent can fetch on demand.

orvexa://models

All available models and their capabilities

orvexa://pricing

Live per-model pricing in USD

orvexa://account/balance

Your current account balance

/.well-known/agent.json — Standard agent discovery manifest for clients that support it.

Prompt Templates

Ready-made prompt templates your agent can invoke to pick the right model for the job.

coding

Coding assistant - routes to a strong code model

long-context

Long-context analysis - routes to a large-context-window model

best-value

Best value - routes to the cheapest model that fits