AtelierEval - Prompt Constraint Coverage - Open-Ended Creation (Novice Prompting): leaderboard

Metric: Prompt checklist coverage (%; share of each task's expert constraint checklist that the written prompt specifies, checked by a GPT-5.4 AtelierJudge; 120 open-ended creation tasks: abstract, narrative requests to turn into an executable prompt; direct natural-language prompting). Source: arxiv.org. Saturation forecast: Estimated already saturated. 8 models tracked.

Top models

#ModelScore
1GPT-5.295.8
2Claude Sonnet 4.594.4
3Qwen 3 VL 235B A22B92.4
4Gemini 3 Pro (Preview)87.8
5GPT-4.186.3
6GPT-4.1 Nano84.3
7Gemini 2.0 Flash82.8
8Qwen 3 VL 8B82.8

Interactive version: theaggregate.ai/benchmark?slug=ateliereval-prompt-constraint-coverage-open-ended-creation-novice-prompting · How It Works · Data refreshed daily, snapshot 2026-09-26.