Eval comparisons by use case

Comparison tables built only from our verified listings.

Coding benchmarks for LLMs, compared — Side-by-side comparison of LLM coding benchmarks — SWE-bench, LiveCodeBench, HumanEval, BigCodeBench, Aider Polyglot and more — with what each tests, license and source.
AI agent and tool-use evals, compared — Compare agent benchmarks for web browsing, computer use, terminals, tool calling and customer-support agents, with what each tests, license and source links.
LLM-as-judge tools, rubrics and eval frameworks — Compare LLM-as-judge rubric libraries, judge templates, judge model APIs and open-source eval frameworks, with who makes each and links to the source.
Image and vision benchmarks, compared — Compare multimodal benchmarks — MMMU, MathVista, MMBench, ChartQA — by what they test, license, last update and source.
RAG and retrieval evals, compared — Compare RAG and retrieval benchmarks and frameworks — Ragas, BEIR, MTEB, CRAG, FRAMES — with what each tests, license and source.