Run 150+ AI apps — image, video, audio, LLMs, 3D and more. Browse, execute, stream results.
https://sh.inference.acAlso advertised: https://api.inference.sh/mcp
Add this server entry to the mcpServers object in your Claude Desktop config, then restart the app.
{
"mcpServers": {
"ac-inference-sh-mcp": {
"url": "https://api.inference.sh/mcp"
}
}
}~/Library/Application Support/Claude/claude_desktop_config.json%APPDATA%\Claude\claude_desktop_config.jsonNo remote HTTP endpoint is advertised. Use the package or stdio setup shown in Install.
inference.sh is an MCP server for Run 150+ AI apps — image, video, audio, LLMs, 3D and more. Browse, execute, stream results.. It supports HTTP transport.
Use the generated config in Install. It points your client at the advertised remote endpoint shown on this page.
Choose the Claude Desktop tab in Install, copy the config, and add it to Claude Desktop's MCP server settings. Endpoint: https://api.inference.sh/mcp.
Choose the Claude Code tab in Install, copy the config, and add it to Claude Code's MCP server settings. Endpoint: https://api.inference.sh/mcp.
Choose the Codex tab in Install, copy the config, and add it to Codex's MCP server settings. Endpoint: https://api.inference.sh/mcp.
Choose the Cursor or VS Code tab in Install, copy the config, and add it to Cursor or VS Code's MCP server settings. Endpoint: https://api.inference.sh/mcp.
inference.sh uses HTTP transport. Endpoint: https://api.inference.sh/mcp.
inference.sh inventory is listed when the MCP endpoint exposes tools, resources, or prompts. Some servers require auth first.
inference.sh does not advertise a verified auth requirement. If discovery fails, it may still need provider login, an API key, a bearer token, or a session header.