InferBench's MCP server lets coding agents run, serve and benchmark local LLMs (text + image, llama.cpp + Stable Diffusion) on your own hardware on demand. Measures real tokens/sec, picks the optimal quant for your GPU, and exposes a 124-model catalog. Local-first, no cloud requi

Installation

⭐ RecommendedC:\Users\<user>\AppData\Local\Programs\InferBench\esources\sidecar\inferbench-backend.exe
C:\Users\<user>\AppData\Local\Programs\InferBench\esources\sidecar\inferbench-backend.exe --mcp
View configuration snippet
{
  "mcpServers": {
    "inferbench": {
      "command": "C:\\Users\\<user>\\AppData\\Local\\Programs\\InferBench\\esources\\sidecar\\inferbench-backend.exe",
      "args": [
        "--mcp"
      ]
    }
  }
}

💡 Tip: Make sure you have the appropriate package manager installed. Refer to the official documentation for detailed installation instructions.

Compatibility Matrix

Filter:

Cursor

STDIO
stdio:From mcp.so (unverified)

Claude Desktop

STDIO
stdio:From mcp.so (unverified)

MCP Compatible

STDIO
stdio:From mcp.so (unverified)
Verified - Tested by community
Unverified - Not yet tested
⚡ STDIO🌊 SSE🌐 HTTP

Server Information

License:unknown
Classification:active
Source:Data aggregated from public MCP directories

Quick Stats

Transports:
STDIO
Compatible Clients:3
Install Methods:1

Like this server?

Discover more MCP servers or submit your own to directory.