multimodal benchmark
QVHighlights
QVHighlights is a video moment retrieval benchmark for detecting moments and highlights in videos via natural language queries. Given a query, the model must localize the start and end times of relevant moments in the video, evaluated using metrics such as Recall@1 at a 0.5 IoU threshold.
Updated Aug 7, 2026
Published models3
Registry coverage3
MetricScore
EvidenceB
QVHighlights leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AM Nova 2 LiteAmazon | 77.2% | 100.0% | 3 | C | |
| 2 | AMNova 2 OmniAmazon | 76.7% | 50.0% | 3 | C | |
| 3 | AMNova 2 ProAmazon | 76.7% | 0.0% | 3 | C |
Score distribution
Top published rows on the benchmark's original scale.
QVHighlights highlights
The top published results on this benchmark's own scale.
What is QVHighlights?
Definition and scoring fields from the benchmark registry.
QVHighlights is a video moment retrieval benchmark for detecting moments and highlights in videos via natural language queries. Given a query, the model must localize the start and end times of relevant moments in the video, evaluated using metrics such as Recall@1 at a 0.5 IoU threshold.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- QVHighlights
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- qvhighlights|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about QVHighlights.
Which model scores highest on QVHighlights?
Nova 2 Lite is currently ranked first with 77.2%.
What does QVHighlights measure?
QVHighlights is a video moment retrieval benchmark for detecting moments and highlights in videos via natural language queries. Given a query, the model must localize the start and end times of relevant moments in the video, evaluated using metrics such as Recall@1 at a 0.5 IoU threshold.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
3 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.