AtelierEval - Prompt Quality (Novice Prompting): leaderboard
Metric: Judge score (1-5; mean of four prompt-level Likert dimensions from a memory-augmented GPT-5.4 judge calibrated on expert-scored exemplars, mean of the three task categories; direct natural-language prompting). Source: arxiv.org. Saturation forecast: Around March 2027. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 | 4.05 |
| 2 | Gemini 3 Pro (Preview) | 3.81 |
| 3 | Qwen 3 VL 235B A22B | 3.65 |
| 4 | GPT-4.1 | 3.46 |
| 5 | Claude Sonnet 4.5 | 3.45 |
| 6 | GPT-4.1 Nano | 3.27 |
| 7 | Qwen 3 VL 8B | 3.15 |
| 8 | Gemini 2.0 Flash | 2.96 |
Interactive version: theaggregate.ai/benchmark?slug=ateliereval-prompt-quality-novice-prompting · How It Works · Data refreshed daily, snapshot 2026-09-26.