# Video Extrapolation on 5 long descriptive prompts (3x / 249f) Wan 2.1 1.3B (train-free extrapolation)

Source: [Video Extrapolation on 5 long descriptive prompts (3x / 249f) Wan 2.1 1.3B (train-free extrapolation)](<https://www.sota2.com/research/sota/video-extrapolation-on-5-long-descriptive-prompts-3x-249f-wan-2-1-1-3b-train-free-extrapolation>)

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

Data updated: 2026-06-01T08:54:42.659Z

Evaluation of video extrapolation models using the specific configuration '5 long descriptive prompts (3x / 249f) Wan 2.1 1.3B' under train-free extrapolation settings. Metrics include VQeval, latency, and speedup.

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 | Video Extrapolation |
| Dataset (including any stated split) | 5 long descriptive prompts (3x / 249f) Wan 2.1 1.3B (train-free extrapolation) |
| Metric | VQeval |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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

- [VQeval](<https://www.sota2.com/research/sota/video-extrapolation-on-5-long-descriptive-prompts-3x-249f-wan-2-1-1-3b-train-free-extrapolation.md?metric=VQeval>): Higher is better; unit: SCALAR.

- [Latency (s)](<https://www.sota2.com/research/sota/video-extrapolation-on-5-long-descriptive-prompts-3x-249f-wan-2-1-1-3b-train-free-extrapolation.md?metric=Latency%20(s)>): Lower is better; unit: TIME.

- [Speedup](<https://www.sota2.com/research/sota/video-extrapolation-on-5-long-descriptive-prompts-3x-249f-wan-2-1-1-3b-train-free-extrapolation.md?metric=Speedup>): Higher is better; unit: SCALAR.

## Recorded results

| Method | VQeval | Result date | Report paper |
| --- | --- | --- | --- |
| LVSA (SDPA) | 62.5 | Not recorded | 6a1d11ecdbd7150506f83e6f |
| LVSA-FI | 62.3 | Not recorded | 6a1d11ecdbd7150506f83e6f |
| UltraViCo | 60.4 | Not recorded | 6a1d11ecdbd7150506f83e6f |
| Dense | 51.1 | Not recorded | 6a1d11ecdbd7150506f83e6f |
| RIFLEx | 51.1 | Not recorded | 6a1d11ecdbd7150506f83e6f |

### LVSA (SDPA)

VQeval: 62.5. Result ID: 6a1d485efbc373592c169a21.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11ecdbd7150506f83e6f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1d11ecdbd7150506f83e6f.md>)

[Implementation](<https://github.com/JiusiServe/LongVideoSparseAttention>)

Recorded evaluation configuration: {"Backbone":"Wan 2.1 1.3B","Attention Backend":"SDPA"}. This record does not assert independent reproduction.

### LVSA-FI

VQeval: 62.3. Result ID: 6a1d485efbc373592c169a22.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11ecdbd7150506f83e6f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1d11ecdbd7150506f83e6f.md>)

[Implementation](<https://github.com/JiusiServe/LongVideoSparseAttention>)

Recorded evaluation configuration: {"Backbone":"Wan 2.1 1.3B","Attention Backend":"FlashInfer"}. This record does not assert independent reproduction.

### UltraViCo

VQeval: 60.4. Result ID: 6a1d485efbc373592c169a20.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11ecdbd7150506f83e6f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6926899ad6b0dd3342930448.md>)

[Implementation](<https://github.com/thu-ml/DiT-Extrapolation>)

Recorded evaluation configuration: {"Backbone":"Wan 2.1 1.3B"}. This record does not assert independent reproduction.

### Dense

VQeval: 51.1. Result ID: 6a1d485efbc373592c169a1e.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11ecdbd7150506f83e6f.md>)

Recorded evaluation configuration: {"Backbone":"Wan 2.1 1.3B"}. This record does not assert independent reproduction.

### RIFLEx

VQeval: 51.1. Result ID: 6a1d485efbc373592c169a1f.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11ecdbd7150506f83e6f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a751e209131cea835c4.md>)

Recorded evaluation configuration: {"Backbone":"Wan 2.1 1.3B"}. This record does not assert independent reproduction.

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