WildVision-Bench — leaderboard

Metric: Overall Reward. Source: huggingface.co. 13 models tracked.

Top models

#ModelScore
1Claude 3.7 Sonnet60.6
2GPT-4.1 (2025-04-14)58.7
3Claude 3.5 Sonnet (20241022)53.2
4InternVL3-38B45.2
5InternVL2.5-78B42.4
6Grok 2 (1212)40
7InternVL3-78B39.6
8InternVL3-8B36.5
9Gemini 2.0 Flash35.5
10Gemini 2.5 Flash32.4
11Llama 4 Scout Instruct31.5
12Pixtral-12B11.8
13Llama 3.2 11B Instruct1.9

Interactive version: aibenchmarks.dev/benchmark?slug=wildvision-bench · How the rankings work · Data refreshed daily, snapshot 2026-07-13.