GigaBench
Judge Your Models Like A Pro.
Suites of Trials with adversarial Judges and a Lead-ready Scorecard by task type.
GigaBench — Scorecard
Suites of Trials with adversarial Judges and a Lead-ready Scorecard by task type.
GigaBench — ScorecardRun models and roles on a task.
Exact, rubric, visual, and more.
The cheapest capable agent.