AICA-Bench: leaderboard

AICA-Bench (2026) tests affective image content analysis by vision-language models: recognizing expressed and evoked emotions, explaining why an image evokes an emotion, and writing emotion-congruent descriptions. The score is the mean of the three task scores in percent, higher is better.

Metric: Overall average on a percent scale: unweighted mean of the emotion understanding average (weighted F1 of basic and CoT prompting), the emotion reasoning score and the emotion-guided content generation score; reasoning and generation are 1-5 ratings rescaled to 20-100, so the floor is 40/3, over AICA-Bench (8,086 affective images from nine public emotion datasets with GPT-4o-generated instructions; closed-source models through their APIs at standard settings, open models on A100 GPUs); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 23 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Pro73.49#276
2Qwen-VL-Max (AICA-Bench checkpoint unspecified)72.9
3GPT-4o72.82#751
4Gemini 2.5 Flash71.14#463
5GPT-4o Mini70.81#1667
6Qwen-VL-Plus (AICA-Bench checkpoint unspecified)67.73
7Gemini 2.0 Flash67.68#723
8Qwen 2.5 VL 7B Instruct65.78#1902
9Ovis2-16B64.78#1594
10Ovis2-8B64.29#2047
11InternVL3-14B-Instruct62.41#1801
12InternVL3-8B-Instruct62.35#2872
13InternVL2.5-8B62.27#2370
14MiniCPM-o-2.661.94#2423
15Qwen 2.5 VL 3B Instruct61.81#2713

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=aica-bench · How It Works · Data refreshed daily, snapshot 2026-11-21.