Towards Evaluation of Implicit Software World Models in Coding LLMs: leaderboard
Metric: F1 (%). Source: benchmarklist.com. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 0.84 |
| 2 | GPT-OSS-120B | 0.6 |
| 3 | Claude Sonnet 4.6 | 0.55 |
| 4 | GPT-5 Mini | 0.5 |
| 5 | Claude Opus 4.7 | 0.49 |
| 6 | Qwen 3.5 397B A17B | 0.45 |
| 7 | GPT-5.2 | 0.39 |
| 8 | GPT-5.4 | 0.38 |
| 9 | Claude Haiku 4.5 | 0.28 |
| 10 | Qwen 3 235B A22B | 0.14 |
| 11 | Qwen 3 30B A3B | 0.05 |
Interactive version: theaggregate.ai/benchmark?slug=towards-evaluation-of-implicit-software-world-models-in-coding-llms · How It Works · Data refreshed daily, snapshot 2026-09-19.