all the models — AI benchmark observatory
← Benchmarks

VITA-Bench

VITA-Bench evaluates AI agents on real-world virtual task automation, measuring their ability to complete complex multi-step tasks in simulated environments.

id vita-bench · max 1 · 11 models reported

#ModelScore

No scores for this benchmark yet.

VITA-Bench Leaderboard · all the models