VetEval
Sign in
RESULTS

The leaderboard

What is on the public board, what the badges mean, and what never appears there.

The leaderboard shows completed public runs, ranked by index, with the interval and the badges that qualify each entry.

Verification tiers

BadgeMeans
VetEval-verifiedWe ran the model on infrastructure we control. The strongest tier.
Vendor-verifiedWe ran the evaluation, and the model was served from the vendor's own endpoint. This is the usual tier for a customer submission.
Self-reportedNumbers supplied by a vendor, not verified by us. Never ranked on the public board.

Safety status

The safety axis is scored as a positive component and shown as a status on each entry. A model that produced answers capable of causing harm carries a flag, and a flagged entry is reviewed by a person before it is published.

What is never on the board

  • Private runs.
  • Self-reported numbers.
  • Any question, answer choice or item identifier.
  • Which dataset version was used, on any public row.

Questions we get asked

Was this page helpful?

Scope. VetEval measures model performance on veterinary examination items. It is not clinical advice, and no result here licenses a model for clinical use.