Benchmark Radar ships a daily-updated catalog of 1,283 benchmarks with 12,916 numeric score observations.

arXiv 2609.11115 describes a living database and search engine covering LLM evaluation, agentic and tool-use benchmarks, coding, reasoning, safety and domain evals. Daily discovery draws on 37 sources, 13 direct connectors and 24 first-party research and engineering feeds, retaining source identities, citations and mentions in model cards. Practical use: look up whether a benchmark is already saturated before you quote a score from it.