DiffCap-Bench - Image Property: leaderboard
DiffCap-Bench (2026) image difference captioning restricted to global image property changes such as resolution or saturation. Macro-averaged F1* in percent, combining the share of true differences described and the share of described differences that are correct, as judged by Gemini 2.5 Pro; higher is better.
Metric: Macro-averaged F1* (%) on global image property changes such as resolution or saturation: harmonic mean of recall* (share of key differences described) and precision* (share of described differences that are correct) on the 1,075 image pairs of DiffCap-Bench (6,713 human-verified difference items), one fixed captioning prompt, open models at temperature 0.1 and proprietary models at their defaults, a Gemini 2.5 Pro judge matching each described difference against the key difference list; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 14 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-5.2 | 89.1 | #190 | |
| 2 | Gemini 3 Pro | 87.8 | #146 | |
| 3 | Kimi K2.5 | Non-reasoning | 87.2 | #272 |
| 4 | Qwen 3 VL Plus | 85.9 | #514 | |
| 5 | Seed1.6-VL | Thinking | 83.6 | #389 |
| 6 | Qwen 3 VL 32B Instruct | 82.9 | #607 | |
| 7 | InternVL3.5-38B | 78.9 | #1026 | |
| 8 | InternVL3.5-8B | 77.2 | #1618 | |
| 9 | Qwen 3 VL 8B Instruct | 76.2 | #1043 | |
| 10 | Qwen 3 VL 32B (Thinking) | 75.9 | #577 | |
| 11 | MiMo-VL-7B-RL-2508 | 75.8 | #829 | |
| 12 | Step3 VL 10B | 75.7 | #1058 | |
| 13 | Qwen 3 VL 8B (Thinking) | 74.1 | #924 | |
| 14 | Grok 4.20 | 72.7 | #307 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=diffcap-bench-image-property · How It Works · Data refreshed daily, snapshot 2026-11-22.