# Anomaly Discrimination on MVTec-AD

Source: [Anomaly Discrimination on MVTec-AD](<https://www.sota2.com/research/sota/anomaly-discrimination-on-mvtec-ad>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:30:16.600Z

Evaluation of image-level anomaly discrimination performance on the MVTec AD dataset, reporting classification accuracy.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Anomaly Discrimination |
| Dataset (including any stated split) | MVTec-AD |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Accuracy](<https://www.sota2.com/research/sota/anomaly-discrimination-on-mvtec-ad.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| LLaVA-One Vision | 94.09 | Not recorded | 68fa76b41e209131cea6e906 |
| Qwen2-VL (+Echo) | 89.65 | Not recorded | 68fa76b41e209131cea6e906 |
| Qwen2-VL | 82.26 | Not recorded | 68fa76b41e209131cea6e906 |
| LLaVA-NeXT | 78.42 | Not recorded | 68fa76b41e209131cea6e906 |
| InternVL2 | 76.88 | Not recorded | 68fa76b41e209131cea6e906 |
| Qwen2-VL | 73.21 | Not recorded | 68fa76b41e209131cea6e906 |
| MiniCPM-V2.6 | 72.5 | Not recorded | 68fa76b41e209131cea6e906 |
| InternVL2 | 70.96 | Not recorded | 68fa76b41e209131cea6e906 |
| Random Chance | 50 | Not recorded | 68fa76b41e209131cea6e906 |

### LLaVA-One Vision

Accuracy: 94.09. Result ID: 69944ad93a6c3148491c4e27.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa756f1e209131cea67e58.md>)

[Implementation](<https://llava-vl.github.io/blog/llava-onevision>)

Recorded evaluation configuration: {"Scale":"7B"}. This record does not assert independent reproduction.

### Qwen2-VL (+Echo)

Accuracy: 89.65. Result ID: 69944ad93a6c3148491c4e2a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

Recorded evaluation configuration: {"Scale":"7B"}. This record does not assert independent reproduction.

### Qwen2-VL

Accuracy: 82.26. Result ID: 69944ad93a6c3148491c4e29.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75151e209131cea65976.md>)

[Implementation](<https://github.com/QwenLM/Qwen2-VL>)

Recorded evaluation configuration: {"Scale":"7B"}. This record does not assert independent reproduction.

### LLaVA-NeXT

Accuracy: 78.42. Result ID: 69944ad93a6c3148491c4e26.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74451e209131cea60aa5.md>)

[Implementation](<https://github.com/LLaVA-VL/LLaVA-NeXT>)

Recorded evaluation configuration: {"Scale":"7B"}. This record does not assert independent reproduction.

### InternVL2

Accuracy: 76.88. Result ID: 69944ad93a6c3148491c4e24.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73081e209131cea59f8b.md>)

[Implementation](<https://github.com/OpenGVLab/InternVL>)

Recorded evaluation configuration: {"Scale":"8B"}. This record does not assert independent reproduction.

### Qwen2-VL

Accuracy: 73.21. Result ID: 69944ad93a6c3148491c4e28.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75151e209131cea65976.md>)

[Implementation](<https://github.com/QwenLM/Qwen2-VL>)

Recorded evaluation configuration: {"Scale":"2B"}. This record does not assert independent reproduction.

### MiniCPM-V2.6

Accuracy: 72.5. Result ID: 69944ad93a6c3148491c4e25.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74681e209131cea6183d.md>)

[Implementation](<https://github.com/OpenBMB/MiniCPM-V>)

Recorded evaluation configuration: {"Scale":"8B"}. This record does not assert independent reproduction.

### InternVL2

Accuracy: 70.96. Result ID: 69944ad93a6c3148491c4e23.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73081e209131cea59f8b.md>)

[Implementation](<https://github.com/OpenGVLab/InternVL>)

Recorded evaluation configuration: {"Scale":"4B"}. This record does not assert independent reproduction.

### Random Chance

Accuracy: 50. Result ID: 69944ad93a6c3148491c4e22.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76b41e209131cea6e906.md>)

Recorded evaluation configuration: {"Scale":null}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.