- Conduid
- Marketplace
- #benchmark
MCP servers tagged benchmark
169 live MCP servers tagged "benchmark", ranked by trust score. Tags come from package metadata and repository topics, so this list covers servers across every category that work with benchmark.
Jaeger MCP Bench
A/B benchmark of MCP tool output formats for Jaeger SPM metrics. 72 randomized trials, Claude + Gemini. Built for jaegertracing/jaeger#8409.
Productagent Evalbench
A local and optional-real-provider Agent evaluation bench for RAG, Memory, Skills, MCP-style tools, tracing, model benchmarking, and cost-controlled evaluation.
Witness Graph
Witness Graph: a falsifiable agent harness where every claim must cite an evidence node in an append-only ledger. Metric Truth Investigator benchmark + scorer.
State Of MCP Security
Reproducible security audit of MCP servers — 75% of scanned servers exposed a medium+ hardening issue to AI agents. Round 1: official reference + popular hosted servers.
Toolsweep
Which tool-schema decision is costing you accuracy? Controlled factorial ablation for LLM tool APIs.
Swe Gym
Reproducible RL environments for evaluating coding agents on real software-engineering tasks through MCP tools. Deterministic verification, golden reference solutions, and CI that…
Answerpath Geo
Privacy-first SEO, GEO & AEO question discovery and search intent mining engine. Categorizes user demand and powers AI visibility benchmarks.
Sol AI
Open agent-reputation infrastructure on Solana: bonded attestations, decayed scores, MCP tools, and a gaming-resistance benchmark
Winagent Lite
Local vision-driven Windows GUI agent with a reproducible benchmark
Forsee MCP
MCP server for the Forsee AI Forum: browse and answer posts, and take a free six-tier capability ladder. Dependency-free stdio server.
Ungrounded
Measure how often your LLM agent reaches for the wrong tool when it can't ground an entity in a request.
LLM Inference Lab
Professional open-source LLM inference benchmarking and optimization toolkit: Transformers, vLLM, llama.cpp, quantization analysis, and GPU profiling.
Honeybadger
formal VM benchmark and inspectable reasoning runtime for testing whether language models can follow machine-like execution semantics on synthetic tasks.
Goku
Goku is an HTTP load testing application written in Rust
Goku
Goku is an HTTP load testing application written in Rust
Awesome AI Gateway
⚡ Awesome AI Gateway — pick an AI gateway from 160+ (LiteLLM, OpenRouter, Portkey, Kong, Higress, new-api, Bifrost) by cost, security & compliance, see what consolidated in 2026,…
Arena
Reliability leaderboard for MCP servers. Scores each server 0-10000 on conformance, tool reliability, agent-task success, hygiene, and honesty.
Defense Bench Py
Score Python MCP defences on mcp-defense-bench: same metrics as the JS leaderboard, plus a pytest regression gate.
Betterthanjson
No description published.