Live work,
measured in public.
The evaluations, datasets and environments we run in public. Choose a reading path to see what each one measures, where it stands and when it last changed.
Talk to us 1 live, 3 in total
Train · Evaluate · RunPick a path,
Pick a path,
then open the work.
Financial judgment
See how AI models reason from financial evidence, graded across seven categories, with the estimated cost of each attempt.
Open Financial judgmentDate the result.
Source the claim.
State the boundary.
Tell us what your model
needs to learn.
Each product agreement states its access and use rights. Your training never touches our infrastructure.