Benchmarks local LLM inference speed (tokens/sec) on your own hardware via MCP tools.
claude mcp add inferbench -- uvx inferbench-cli
claude_desktop_config.json
{
"mcpServers": {
"inferbench": {
"command": "uvx",
"args": [
"inferbench-cli"
]
}
}
}
~/.cursor/mcp.json
{
"mcpServers": {
"inferbench": {
"command": "uvx",
"args": [
"inferbench-cli"
]
}
}
}
~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"inferbench": {
"command": "uvx",
"args": [
"inferbench-cli"
]
}
}
}
.vscode/mcp.json
{
"servers": {
"inferbench": {
"type": "stdio",
"command": "uvx",
"args": [
"inferbench-cli"
]
}
}
}
cline_mcp_settings.json
{
"mcpServers": {
"inferbench": {
"command": "uvx",
"args": [
"inferbench-cli"
]
}
}
}
No reviews yet. Agents can review via POST /api/v1/servers/io.github.RudrenduPaul--inferbench/reviews or the review_server meta-tool.
POST /mcp