Coding benchmarks
Understand code generation, repository repair and software engineering evaluations before comparing model results.
199 source recordsAI benchmark topics
Explore AI evaluations by capability and domain. Topics can overlap; sharing a topic does not make two scores comparable.
Computer use benchmarks
Read evaluations of GUI grounding, browser actions and complete desktop or mobile workflows.
18 source recordsTools and agents benchmarks
Distinguish correct tool calls from successful multi-step work, and read results as evaluations of a configured system.
301 source recordsMultimodal benchmarks
Find out which combinations of images, text, audio or video a benchmark actually evaluates.
372 source recordsReasoning and math benchmarks
Read reasoning results through the problem type, permitted tools and definition of a correct answer.
674 source recordsLong context benchmarks
Distinguish accepting a long input from reliably retrieving, combining and summarizing information inside it.
81 source recordsLanguage benchmarks
Interpret translation, instruction following and text quality results in the language and task where they were measured.
164 source recordsSafety and reliability benchmarks
Separate harmful compliance, unnecessary refusal and reliability failures when interpreting safety evaluations.
64 source recordsCybersecurity benchmarks
Distinguish CTF solving, vulnerability reproduction, exploit capability and secure coding results.
10 source recordsFinance benchmarks
Understand financial document questions, numerical reasoning and evidence retrieval without confusing them with investment performance.
32 source recordsMedical benchmarks
Distinguish medical knowledge questions from clinical documentation, communication and decision-support evaluations.
56 source recordsScience benchmarks
Separate scientific knowledge, quantitative reasoning and research code when choosing an evaluation.
88 source recordsMAKE SENSE OF THE SCORES
Find the right evaluation for your task
Start with the work you want an AI system to do. Then check the task definition, scoring rule and evaluation setup. Coding, cybersecurity, computer use and specialist-domain benchmarks answer different questions; a score is useful only with its source and context.
Explore the twelve benchmark topics
AI benchmarks and leaderboards are different things
Learn the difference between a benchmark, an evaluation run and a leaderboard, and use each to answer the right question.
How to compare AI benchmark scores without mixing protocols
A practical sequence for checking task versions, metrics, tools, budgets, uncertainty and reporting sources.
Choose benchmarks for coding, cybersecurity and computer use agents
Match your agent's job to a verifiable outcome, then build a small evaluation portfolio that covers its distinct failure modes.