What it does
InferBench exposes MCP tools that measure how fast a large language model runs locally, reporting throughput in tokens per second. The publisher describes it as a way to benchmark inference speed directly on the machine where the model runs, rather than relying on remote services.
It would be reached for when someone wants to compare local model performance on their own hardware, for example to check how a model responds on a given machine before choosing it for development work. The publisher's description covers only the measurement of tokens per second through MCP tools; other details of how it is run are not stated in the available facts.
Safety
Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.
- Runs code or changes an assistant's behaviour
Sources
- Official MCP Registry first seen 13 Nov 2026 · last seen 13 Nov 2026 · its page on the source
io.github.RudrenduPaul/inferbench
Benchmarks local LLM inference speed (tokens/sec) on your own hardware via MCP tools.
v0.1.7As shown on registry.modelcontextprotocol.io when we last read it