Benchmark
TVBench
What is TVBench?
TVBench is a temporal video understanding benchmark evaluating reasoning over actions, events, and temporal dynamics in videos.
- Modality
- multimodal
- Evaluates
- multimodal, vision
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 3
Source provenance
- Original evidence https://api.zeroeval.com/leaderboard/benchmarks/tvbench?top_n=500
TVBench results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Qwen3.7-Plus | Qwen | 0.782 | 2026-05-31 | evidence |
| Seed 2.1 Pro | ByteDance | 0.805 | 2026-06-24 | evidence |
| Seed 2.1 Turbo | ByteDance | 0.772 | 2026-06-24 | 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.