“research”

Simulated customer-service benchmark where an agent follows domain policies and uses tools while a simulated user also acts.

Benchmark and framework that tests models and agents on tasks using real-world MCP servers.

1,266 hard-to-find questions with short answers for testing how persistently agents can search the web.

Self-hosted web environment with realistic sites (shopping, forums, code hosting, maps) for testing web agents.

Continuously refreshed coding benchmark built from new contest problems to limit training-data contamination.