OpenMayhem MCP
MCP serverMediaUse OpenMayhem text, embedding and media models with explicit cost controls.
Available today. Use it from your connected AI after setup.
No other account needed.
Connect ahel once, and every AI you use reads what you have installed.
Then ask your AI: use OpenMayhem MCP
From the project's README
As published by Trac-Systems/openmayhem-mcp in README.md.
OpenMayhem's local stdio bridge forwards MCP messages to the hosted OpenMayhem service. It does not bundle a stale catalog: list_models and get_model read the live catalog, availability, capabilities, limits, routes, and prices from OpenMayhem. Account balance and payment logic remain on the hosted service, and the bridge contains no provider data or credentials.
Install
Download openmayhem-mcp-v1.0.0.mcpb from the v1.0.0 release and install it in an MCPB-compatible client. The client asks for a restricted OpenMayhem API key and stores it as a sensitive setting.
Clients that support remote Streamable HTTP can connect directly:
URL: https://mcp.openmayhem.ai/mcp
Header: X-OpenMayhem-API-Key: your restricted OpenMayhem API key
Create and revoke API keys at openmayhem.ai. Complete setup, tool, payment, and client instructions are at openmayhem.ai/docs/mcp.
Billing
Model discovery, request estimation, token counting, balance checks, job reads, cancellation, and owned-artifact lookup do not spend model credit. Model execution is paid. Every paid tool requires a current estimate and an explicit maximum cost.
The bridge does not calculate charges or handle payment information. Requests use the existing OpenMayhem account, balance, reservation, receipt, and settlement paths.
Requirements
- Node.js 22 or newer
- A restricted OpenMayhem API key
- HTTPS access to
mcp.openmayhem.ai
License
MIT
Advanced
- Delivery
- openmayhem MCP server → your ahel gateway (mcp.ahel.ai) → every connected AI client.
- Catalog kind
- mcp-server
- Gateway key
ai-openmayhem-openmayhem- Source
- github.com/Trac-Systems/openmayhem-mcp
- Hosted endpoint
https://mcp.openmayhem.ai/mcp