WorldModelBench physics componentBrowse 296

WorldModelBench physics component

Can the model plan, use tools, and finish a multi-step task?

Version not specifiedExact reported variant
Agents11 ranked models11 reported values1 reportsHigher is better

Top rankings

One row per model, using its best reported score across effort settings.

11
RankModelBest scoreBest reported setting
1CosmosNVIDIA1.00Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
2KlingExternal baseline1.00Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
3LTX-2External baseline1.00Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
4Qwen-RobotWorldQwen1.00Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
5VidarExternal baseline1.00Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
6GigaWorldExternal baseline0.99Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
7LVPExternal baseline0.99Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
8Sora2OpenAI0.99Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
9Veo 3Google0.99Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
10Wan2.6External baseline0.99Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source
11WowExternal baseline0.99Reported configurationQwen report1 value · 1 reportJun 15, 2026 · source

Effort curve

Every sourced cost-linked effort value for this exact version. Lines connect complete sweeps only.

0
No cost-linked effort sweep for this version.
Definition and comparison boundarypublic methodology

Can the model plan, use tools, and finish a multi-step task?

Agentic execution across tools, environments, or long-running workflows. Higher is better. This benchmark reports points rather than percent correct.

Harness, tools, prompts, attempt count, and benchmark version can materially change the result. Compare within one reporting context.

WorldModelBench physical-reasoning component scores; lower-level component names follow the source table.

Method / source