Video Arena — rank the best AI video models by human preference
Video Arena is a blind, human-preference benchmark for AI video generation models. Each matchup is pre-generated from an operator-curated prompt and source image. Two anonymous models receive the same input, each produces a short clip, and visitors vote for the more convincing result; the models are revealed only after you vote. Anonymous votes are recorded for internal benchmark analysis.
How it works
- Open a random curated matchup.
- Two anonymous models have generated clips from the same input.
- Vote blind for the better result, or call it a tie.
- Identities are revealed after the vote; the result is recorded for benchmark analysis.
What it ranks
Frontier image-to-video models — from labs such as Google (Veo), Kuaishou (Kling), OpenAI (Sora), and ByteDance (Seedance) — scored purely by what people prefer, not by automated metrics that correlate poorly with perceived quality.
Open Video Arena and vote on a blind comparison.