Imaging-101 - Function-Level Preprocessing: leaderboard
A function-level module of Imaging-101 (2026): the percent of its 57 computational imaging tasks whose preprocessing module passes all of its unit tests within 10 reflection rounds. Higher is better.
Metric: Module pass rate (%): share of the 57 tasks whose preprocessing module passes all of its unit tests within 10 reflection rounds; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 78.9 | #94 |
| 2 | Claude Opus 4.6 | 75.4 | #102 |
| 3 | DeepSeek V3 | 75.4 | #716 |
| 4 | GPT-5.4 | 73.7 | #123 |
| 5 | Qwen 3.6 Plus | 73.7 | #222 |
| 6 | Kimi K2.5 | 61.4 | #256 |
| 7 | GLM-5 | 45.6 | #260 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=imaging-101-function-level-preprocessing · How It Works · Data refreshed daily, snapshot 2026-11-13.