BoxComm (Punch Events): leaderboard
BoxComm boxing commentary generation with the video plus the most recent automatically detected punch events supplied as text (the authors EIC-Gen input). The score is the percentage of generated sentences that GPT-5.2 judges consistent with the reference commentary, higher is better.
Metric: Content consistency accuracy (%) over all requested categories; input video plus the 16 most recent detected punch events as text (the authors EIC-Gen input); category-conditioned commentary generation on BoxComm-Eval (the last 40 of 445 World Boxing Championship broadcast matches, 5.6K professional commentary sentences): from the 4 seconds of video before the target moment, the 8 previous commentary sentences and a requested category (play-by-play, tactical or contextual) the model writes one commentary sentence, and GPT-5.2 judges whether it conveys the core action or tactical meaning of the reference sentence; zero-shot; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 3 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 VL 8B Instruct | 23.4 | #1029 |
| 2 | LLaVA-OV-7B (BoxComm checkpoint unspecified) | 21.3 | |
| 3 | GPT-4o Mini | 19.5 | #1667 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=boxcomm-punch-events · How It Works · Data refreshed daily, snapshot 2026-11-21.