Video generation
Last updated: March 11, 2025Our video generation leaderboard evaluates AI models on their ability to generate high-quality videos from textual descriptions. We assess factors such as visual quality, adherence to the given text, and creativity.
Rank | Model | Elo rating | TrueSkill rating |
|---|---|---|---|
 1 | Runway Gen 3 | 1152.5 | 1089.64 |
2 | Luma Ray 2 | 1099.75 | 1115.1 |
3 | Tencent | 1035.55 | 1014.48 |
4 | Pika 1.5 | 860.54 | 881.36 |
5 | Luma Dreamachine | 851.66 | 878.37 |
What is “Elo rating”?
This is a dynamic rating system used in competitive games to rank players. In this context, it's applied to models. Higher Elo ratings indicate better performance based on head-to-head rankings. The Elo system adjusts ratings based on how well models perform against each other, and the K-factor (32) determines how much the rating changes after each match.
Human preference evaluation
Diverse pool of US-based Alignerrs, including generalists and creative artists
Consensus of three Alignerrs per task
Standardized instructions and ontology for consistent evaluations
Carefully curated prompt generation process, balancing creativity and clarity
Consistency with prompt
Artifacts and errors
Realistic
Description:
Evaluate how well the input prompt is represented in the output video content.
Options:
High
Medium
Low
