MeduxMedux
Overview / MCP

Start with Medux MCP

Connect your agent to the Medux-hosted remote MCP service to call media tools with the same account, API key, tasks, and credits system as the public API.

Browse API Reference

How to start

1

Sign in and get your Medux API key.

2

Point your agent's MCP client at the Medux-hosted remote MCP service — one fixed endpoint, the same for every capability. Claude Code and Codex can add it from a command or a config snippet; any other MCP client can use the same shape as JSON.

claude mcp add --transport http medux https://medux-server-api-dev.lingverse.co --header "Authorization: Bearer sk-medux-..."
3

Open Playground, switch to For AI Agents, and read the MCP tab to get the mcp_tool_id for the capability you want to call, then call that tool from your agent.

4

Track task status, results, and credits usage with the same Medux account system.

What this means

  • Medux hosts the remote MCP service instead of requiring users to deploy a local server.
  • Every executable Playground capability maps to a stable mcp_tool_id with links back to API Reference and Playground.
  • Remote MCP calls follow the same plan limits, rate limits, credits estimates, holds, and settlement rules as public API calls.

What you get

Official hosted remote MCP service.

Stable tool identifiers, unified auth, task status, and result structure.

Shared credits, task, asset, and error semantics with the public API.