Benchmark
CharadesSTA Benchmark
What is CharadesSTA?
Charades-STA is a benchmark dataset for temporal activity localization via language queries, extending the Charades dataset with sentence temporal annotations…
- Released
- 2017-05-05
- Modality
- multimodal
- Evaluates
- multimodal, language, video, vision
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 12
Source provenance
- Original evidence https://api.zeroeval.com/leaderboard/benchmarks/charadessta?top_n=500
CharadesSTA results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Qwen2.5 VL 32B Instruct | Qwen | 0.542 | 2025-02-28 | evidence |
| Qwen2.5 VL 7B Instruct | Qwen | 0.436 | 2025-01-26 | evidence |
| Qwen3 VL 235B A22B Instruct | Qwen | 0.648 | 2025-09-22 | evidence |
| Qwen3 VL 235B A22B Thinking | Qwen | 0.635 | 2025-09-22 | evidence |
| Qwen3 VL 30B A3B Instruct | Qwen | 0.635 | 2025-09-22 | evidence |
| Qwen3 VL 30B A3B Thinking | Qwen | 0.627 | 2025-09-22 | evidence |
| Qwen3 VL 32B Instruct | Qwen | 0.612 | 2025-09-22 | evidence |
| Qwen3 VL 32B Thinking | Qwen | 0.628 | 2025-09-22 | evidence |
| Qwen3 VL 4B Instruct | Qwen | 0.555 | 2025-09-22 | evidence |
| Qwen3 VL 4B Thinking | Qwen | 0.59 | 2025-09-22 | evidence |
| Qwen3 VL 8B Instruct | Qwen | 0.56 | 2025-09-22 | evidence |
| Qwen3 VL 8B Thinking | Qwen | 0.599 | 2025-09-22 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.