Benchmarks

Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.

309,131 indexed

Task	Dataset	Domain	Results	Last Update
Multimodal Understanding	MMBench (dev)	multimodal	73	Jun 4, 2026
Long-context language understanding	LongBench 1.0 (test)	llm chat instructionnlp	73	Jun 9, 2026
Multiple-choice Question Answering	MMLU 5-shot	nlp	73	Apr 2, 2026
Moment Retrieval	QVHighlights (val)	computer visionnlp