FoodSense - Sound: leaderboard

FoodSense (2026) cross-sensory inference for sound: from a food photo the model predicts a 1-5 rating, compared with the mean human rating on 438 test images by Lin's concordance correlation coefficient (-1 to 1, shown times 100); higher is better.

Metric: Sound: Lin's concordance correlation coefficient between the model's predicted 1-5 rating and the mean human rating (about 22 annotators per image) on FoodSense's 438 held-out test food images, zero-shot with the authors' sensory-assessment prompt, ratings parsed by regex; range -1 to 1, shown times 100; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 5 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1LLaVA-v1.6-34B15.9#5950
2Qwen 2.5 VL 32B Instruct14#1062
3InternVL2.5-26B13.4#2156
4Gemma 3 27B (IT)5.3#1324
5food-Llama-3.2-11B-Vision-Instruct4

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

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