# Image Classification on ImageNet-1k (Accuracy, Model Size, Latency)

Source: [Image Classification on ImageNet-1k (Accuracy, Model Size, Latency)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-accuracy-model-size-latency>)

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

Data updated: 2026-05-26T09:08:56.956Z

Evaluation of Image Classification performance on the ImageNet-1k dataset, reporting standard accuracy alongside model size (MB) and inference latency (Dec. s).

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 |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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/image-classification-on-imagenet-1k-accuracy-model-size-latency.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [Model Size (MB)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-accuracy-model-size-latency.md?metric=Model%20Size%20(MB)>): Lower is better; unit: SIZE.

- [Latency (decisecs)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-accuracy-model-size-latency.md?metric=Latency%20(decisecs)>): Lower is better; unit: TIME.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| MCWC (K = 4) | 81.05 | Not recorded | 6a15293bdbd7150506f7e0cf |
| QAT INT4 | 80.2 | Not recorded | 6a15293bdbd7150506f7e0cf |

### MCWC (K = 4)

Accuracy: 81.05. Result ID: 6a156321fbc373592c0d0654.

[Reporting paper](<https://www.sota2.com/research/paper/6a15293bdbd7150506f7e0cf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a15293bdbd7150506f7e0cf.md>)

Recorded evaluation configuration: {"Model":"Swin-T","bits per model parameter":"4.2"}. This record does not assert independent reproduction.

### QAT INT4

Accuracy: 80.2. Result ID: 6a156321fbc373592c0d0655.

[Reporting paper](<https://www.sota2.com/research/paper/6a15293bdbd7150506f7e0cf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67741e209131cea16685.md>)

Recorded evaluation configuration: {"Model":"Swin-T","bits per model parameter":"4.3"}. This record does not assert independent reproduction.

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