Who leads the VisA benchmark?

simplenet currently leads VisA with a score of 95.5 on auroc.

What is the state-of-the-art score on VisA?

The state-of-the-art result on VisA is 95.5 (auroc), achieved by simplenet as of 2025.

How many models are tracked on VisA?

Codesota tracks 3 models on VisA.

When was the VisA leaderboard last updated?

The VisA leaderboard on Codesota includes results through 2025.

Codesota · Benchmark · VisAHome/Leaderboards/VisA

Unknown

VisA.

Name: VisA Benchmark Results
Creator: Unknown
Published: 2025-01-01
License: https://creativecommons.org/licenses/by/4.0/

10,821 high-resolution images across 12 objects with complex structures. More challenging than MVTec with realistic industrial scenarios.

Paper ↗Leaderboard ↓

§ 01 · Leaderboard

Results by metric.

Only 3 models on this benchmark

Help build the community leaderboard — submit your model results.

Found a wrong score or missing run?

Use row edits to send a sourced correction into moderation.

Add / edit result ↗Report issue ↗

auroc

Auroc is the reported evaluation metric for VisA. Codesota tracks published model scores on this metric so readers can compare state-of-the-art results across sources and model families.

Higher is better

Trust tiers for aurocverifiedpapervendorcommunityunverified

Rank	Model	Trust	Score	Year	Links	Fix
01	simplenet Strong generalization to VisA.	paper	95.5	2025	Source ↗	Looks wrong?
02	efficientad Fast inference on VisA.	paper	94.8	2025	Source ↗	Looks wrong?
03	patchcore VisA is more challenging than MVTec.	paper	92.1	2025	Source ↗	Looks wrong?

§ 04 · Submit a result

Add to the leaderboard.

← Back to Leaderboards