Do Enterprise Systems Need Learned World Model: leaderboard
World models enable agents to anticipate the effects of their actions by internalizing environment dynamics. In enterprise systems, however, these dynamics are often defined by tenant-specific business logic that varies across deployments and evolves over time, making models trained on historical transitions brittle under deployment shift. We ask a question the world-models literature has not addr
Metric: CascadeBench IoU w/ BR (self-reported). Source: benchmarklist.com. Saturation forecast: Around August 2027. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 41.78 | #98 |
| 2 | Gemini 3 Pro | 41.36 | #80 |
| 3 | Claude Opus 4.6 | 40.46 | #63 |
| 4 | Qwen 3.5 27B | 40.39 | #187 |
| 5 | Qwen 3.6 27B | 38.98 | #170 |
| 6 | Claude Sonnet 4.6 | 38.15 | #90 |
| 7 | Gemma 4 31B | 37.99 | #200 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=do-enterprise-systems-need-learned-world-model · How It Works · Data refreshed daily, snapshot 2026-10-20.