# Image Classification on ImageNet 1K (test) (Top-1 Accuracy and GAIN)

Source: [Image Classification on ImageNet 1K (test) (Top-1 Accuracy and GAIN)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-test-top-1-accuracy-and-gain>)

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

Data updated: 2026-02-26T06:06:49.718Z

Evaluation of Image Classification performance on the test split of the ImageNet 1k dataset, reporting Top-1 Accuracy and the GAIN metric.

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 | Image Classification |
| Dataset (including any stated split) | ImageNet 1K (test) |
| Metric | Top-1 Acc |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 28 of 28 |
| 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

- [Top-1 Acc](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-test-top-1-accuracy-and-gain.md?metric=Top-1%20Acc>): Higher is better; unit: PERCENT.

- [GAIN](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-test-top-1-accuracy-and-gain.md?metric=GAIN>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Top-1 Acc | Result date | Report paper |
| --- | --- | --- | --- |
| DINOv3 | 87.1 | Not recorded | 68fa7be61e209131cea8b292 |
| SigLIP2 | 87 | Not recorded | 68fa7be61e209131cea8b292 |
| SigLIP | 86.1 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 85.9 | Not recorded | 68fa7be61e209131cea8b292 |
| AIMv2 | 85.9 | Not recorded | 68fa7be61e209131cea8b292 |
| DINOv2 | 85.6 | Not recorded | 68fa7be61e209131cea8b292 |
| DINOv3 | 84.4 | Not recorded | 68fa7be61e209131cea8b292 |
| Franca | 84.3 | Not recorded | 68fa7be61e209131cea8b292 |
| DINOv2 | 84 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 83.6 | Not recorded | 68fa7be61e209131cea8b292 |
| CAPI | 83.6 | Not recorded | 68fa7be61e209131cea8b292 |
| CLIP | 83.4 | Not recorded | 68fa7be61e209131cea8b292 |
| BEiTv2 | 81.7 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 80.6 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 79.3 | Not recorded | 68fa7be61e209131cea8b292 |
| iBOT | 79.2 | Not recorded | 68fa7be61e209131cea8b292 |
| DINO | 77.8 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 76 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 75.6 | Not recorded | 68fa7be61e209131cea8b292 |
| BYOL | 75.1 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 67.7 | Not recorded | 68fa7be61e209131cea8b292 |
| SimMIM | 65.1 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 64.6 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 58.2 | Not recorded | 68fa7be61e209131cea8b292 |
| DiT | 57 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 47.4 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 46.1 | Not recorded | 68fa7be61e209131cea8b292 |
| MAE | 26.7 | Not recorded | 68fa7be61e209131cea8b292 |

### DINOv3

Top-1 Acc: 87.1. Result ID: 699ae28c5698665ec4728501.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a931e209131cea83fc5.md>)

Recorded evaluation configuration: {"ARCH":"ViT-L/16","PRE-TRAINING":"LVD-1689M","Evaluation Protocol":"EP","# PAR.":"2.1M"}. This record does not assert independent reproduction.

### SigLIP2

Top-1 Acc: 87. Result ID: 699ae28c5698665ec4728504.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa770a1e209131cea70ad4.md>)

Recorded evaluation configuration: {"ARCH":"ViT-L/16","PRE-TRAINING":"WebLI","Evaluation Protocol":"EP","# PAR.":"2.1M"}. This record does not assert independent reproduction.

### SigLIP

Top-1 Acc: 86.1. Result ID: 699ae28c5698665ec4728503.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70ae1e209131cea4c7a4.md>)

Recorded evaluation configuration: {"ARCH":"ViT-L/16","PRE-TRAINING":"WebLI","Evaluation Protocol":"EP","# PAR.":"2.1M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 85.9. Result ID: 699ae28c5698665ec47284f6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-L/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"FT","# PAR.":"304M"}. This record does not assert independent reproduction.

### AIMv2

Top-1 Acc: 85.9. Result ID: 699ae28c5698665ec4728506.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75db1e209131cea6a431.md>)

[Implementation](<https://github.com/apple/ml-aim>)

Recorded evaluation configuration: {"ARCH":"ViT-L/14","PRE-TRAINING":"custom\*","Evaluation Protocol":"EP","# PAR.":"2.1M"}. This record does not assert independent reproduction.

### DINOv2

Top-1 Acc: 85.6. Result ID: 699ae28c5698665ec47284fe.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72181e209131cea546c3.md>)

[Implementation](<https://github.com/facebookresearch/dinov2>)

Recorded evaluation configuration: {"ARCH":"ViT-L/14","PRE-TRAINING":"LVD-142M","Evaluation Protocol":"EP","# PAR.":"2.1M"}. This record does not assert independent reproduction.

### DINOv3

Top-1 Acc: 84.4. Result ID: 699ae28c5698665ec4728500.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a931e209131cea83fc5.md>)

Recorded evaluation configuration: {"ARCH":"ViT-B/16","PRE-TRAINING":"LVD-1689M","Evaluation Protocol":"EP","# PAR.":"1.4M"}. This record does not assert independent reproduction.

### Franca

Top-1 Acc: 84.3. Result ID: 699ae28c5698665ec47284ff.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a611e209131cea82bf4.md>)

Recorded evaluation configuration: {"ARCH":"ViT-L/14","PRE-TRAINING":"IN-21k","Evaluation Protocol":"EP","# PAR.":"2.1M"}. This record does not assert independent reproduction.

### DINOv2

Top-1 Acc: 84. Result ID: 699ae28c5698665ec47284fd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72181e209131cea546c3.md>)

[Implementation](<https://github.com/facebookresearch/dinov2>)

Recorded evaluation configuration: {"ARCH":"ViT-B/14","PRE-TRAINING":"LVD-142M","Evaluation Protocol":"EP","# PAR.":"1.4M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 83.6. Result ID: 699ae28c5698665ec47284f2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-B/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"FT","# PAR.":"87M"}. This record does not assert independent reproduction.

### CAPI

Top-1 Acc: 83.6. Result ID: 699ae28c5698665ec47284f9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79b81e209131cea7f17d.md>)

Recorded evaluation configuration: {"ARCH":"ViT-L/14","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"2.1M"}. This record does not assert independent reproduction.

### CLIP

Top-1 Acc: 83.4. Result ID: 699ae28c5698665ec4728502.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"ARCH":"ViT-L/14","PRE-TRAINING":"WIT","Evaluation Protocol":"EP","# PAR.":"2.1M"}. This record does not assert independent reproduction.

### BEiTv2

Top-1 Acc: 81.7. Result ID: 699ae28c5698665ec47284f7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d8e1e209131cea3a924.md>)

Recorded evaluation configuration: {"ARCH":"ViT-B/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"1.4M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 80.6. Result ID: 699ae28c5698665ec47284ee.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-S/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"FT","# PAR.":"22M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 79.3. Result ID: 699ae28c5698665ec47284f5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-L/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"2.1M"}. This record does not assert independent reproduction.

### iBOT

Top-1 Acc: 79.2. Result ID: 699ae28c5698665ec47284fc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bb51e209131cea2fbe0.md>)

[Implementation](<https://github.com/bytedance/ibot>)

Recorded evaluation configuration: {"ARCH":"ViT-B/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"1.4M"}. This record does not assert independent reproduction.

### DINO

Top-1 Acc: 77.8. Result ID: 699ae28c5698665ec47284fb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a071e209131cea2594c.md>)

[Implementation](<https://github.com/facebookresearch/dino>)

Recorded evaluation configuration: {"ARCH":"ViT-B/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"1.4M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 76. Result ID: 699ae28c5698665ec47284f4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-L/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"LP","# PAR.":"1.0M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 75.6. Result ID: 699ae28c5698665ec47284f1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-B/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"1.4M"}. This record does not assert independent reproduction.

### BYOL

Top-1 Acc: 75.1. Result ID: 699ae28c5698665ec47284fa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa684a1e209131cea1b1e9.md>)

[Implementation](<https://github.com/deepmind/deepmind-research/tree/byol>)

Recorded evaluation configuration: {"ARCH":"RN-50","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"6.3M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 67.7. Result ID: 699ae28c5698665ec47284f0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-B/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"LP","# PAR.":"0.8M"}. This record does not assert independent reproduction.

### SimMIM

Top-1 Acc: 65.1. Result ID: 699ae28c5698665ec47284f8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c5b1e209131cea33891.md>)

[Implementation](<https://github.com/microsoft/SimMIM>)

Recorded evaluation configuration: {"ARCH":"ViT-B/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"1.4M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 64.6. Result ID: 699ae28c5698665ec47284ed.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-S/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"0.5M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 58.2. Result ID: 699ae28c5698665ec47284f3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-L/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"k-NN"}. This record does not assert independent reproduction.

### DiT

Top-1 Acc: 57. Result ID: 699ae28c5698665ec4728505.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ea91e209131cea41328.md>)

Recorded evaluation configuration: {"ARCH":"DiT-XL/2","PRE-TRAINING":"IN-1K","Evaluation Protocol":"EP","# PAR.":"2.5M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 47.4. Result ID: 699ae28c5698665ec47284ec.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-S/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"LP","# PAR.":"0.4M"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 46.1. Result ID: 699ae28c5698665ec47284ef.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-B/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"k-NN"}. This record does not assert independent reproduction.

### MAE

Top-1 Acc: 26.7. Result ID: 699ae28c5698665ec47284eb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b292.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e973.md>)

Recorded evaluation configuration: {"ARCH":"ViT-S/16","PRE-TRAINING":"IN-1K","Evaluation Protocol":"k-NN"}. This record does not assert independent reproduction.

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