MethodologyAnnotator qualityUpdatesHugging Face436.7k downloadsSuggest a model or benchmarkRequest a private benchmark
Benchmarks
Dimensions
License
Models

Scores recompute with the real footage in the field.

Realism v1updated 2026-10-06

World Models Realism: AI vs. real camera footage

Rapidatabenchmark@rapidata.ai

Abstract

This benchmark ranks world and video generation models on whether their videos pass for real camera footage, judged blind against each other and against the real clip each scene was taken from. We took 25 real clips of everyday motion, from a plane landing to coffee being poured, and had every model generate four videos per clip from its opening frame and a one-line description. The rankings are updated continuously from 618,210 responses across 85,098 pairwise human-preference matches. All data and judgments are public; see the methodology for details on evaluation and scoring.

LeaderboardCompare OutputsMatchupsAnalysisPromptsDetails
#
Model
Organization
Score
95% CI
Interval
Responses
Matches
Win rate
#
Model
Score

Table 1. Overall standings, World Models Realism benchmark. Scores from a Bradley–Terry fit rescaled to Elo; intervals are 95% Wald intervals. Click any row to view sample outputs in Figure 1. The real footage is scored alongside the models as a reference and takes no rank.

Methodology notesCompare Outputs
Figure 1. Sample outputs from MiniMax H3 on representative prompts from the fixed set. Outputs shown are those used in matches.
Score vs. cost
1000110012001300$0.01$0.02$0.05$0.1$0.2$0.5cost, USD per video second (log scale)score

Figure 2. Overall Elo of each model against its list price per second of video, on a logarithmic axis. Prices are recorded manually from the providers’ published rates. Error bars are 95% confidence intervals of the score. Models ringed in teal and joined by the step line form the Pareto frontier: no other model has both a lower cost and a higher score. The remaining models are faded. 6 models without a cost in USD per video second are not shown.

Evaluation: Models are compared on the same prompts through pairwise human judgements.
Score: Scores show relative performance within this benchmark. A higher score means a model was more likely to be preferred over the other evaluated models.
Uncertainty: The 95% confidence interval shows the uncertainty around each model’s estimated score.
Views: Results can be explored by photorealism, visual integrity.
Coverage: This benchmark includes a fixed prompt set, evaluated on all ranked models.
Data: Available prompts, model outputs, and evaluation results can be downloaded for further analysis.
Version: Results correspond to Realism v1, last updated on 2026-11-19.

We use crowd intelligence to deliver real, diverse human feedback at scale for model evaluation and post-training.

Benchmarks
Text-to-ImageOutput Diversity (Mode Collapse)Multi-Step Image Editing (Identity Drift)Camera MovementRelative Camera MovementPhysicsPhysicsRealismImage-to-3DSVGLLM-to-3DLLM to AnimationSuggest a model or benchmark
Method
MethodologyData on Hugging Face
Rapidata
rapidata.ai Private benchmarksContactX LinkedIn
Rapidata, benchmark.ai © 2026