BEIR (Benchmarking IR) is a benchmark suite for evaluating information retrieval and vector search systems across multiple tasks and datasets. Useful for comparing vector database performance.

Massive Text Embedding Benchmark leaderboard covering 58 datasets across 112 languages and 8 embedding tasks. Industry-standard benchmark for comparing text embedding models.

Comprehensive benchmark for evaluating long-term memory in chat assistants with 500 manual questions testing information extraction, multi-session reasoning, and temporal reasoning across 115K-1.5M tokens.

Massive Multilingual Text Embedding Benchmark covering over 500 quality-controlled evaluation tasks across 250+ languages, representing the largest multilingual collection of embedding model evaluation tasks.

Billion-scale benchmark dataset containing 128-dimensional SIFT descriptors of one billion images. Widely used standard for evaluating approximate nearest neighbor search algorithms at scale.

An annual competition focused on similarity search and indexing algorithms, including approximate nearest neighbor methods and high-dimensional vector indexing, providing benchmarks and results relevant to vector database research.