Frontier Leaderboards
Legacy Leaderboards
2025 Scale AI. All rights reserved.
VISTA
Visual Language Understanding
Last updated: April 4, 2025
Performance Comparison
1
54.65±1.46
2
51.79±0.63
2
51.66±1.08
2
50.78±0.57
2
50.07±1.14
4
49.59±0.66
5
47.32±1.78
6
48.23±0.70
7
46.97±1.29
8
45.50±1.20
8
45.34±0.91
9
45.25±0.40
11
43.25±1.26
13
43.02±1.14
13
42.11±1.39
15
41.14±0.58
15
39.95±0.80
16
39.85±0.71
17
Claude 3.5 Sonnet (October 2024)
38.72±0.51
19
Claude 3.5 Sonnet (June 2024)
38.37±0.70
19
38.33±0.55
19
ChatGPT-4o-latest (November 2024)
37.99±0.48
19
Gemini 1.5 Pro
37.07±1.34
24
GPT-4o (August 2024)
34.94±0.23
24
34.59±1.12
24
Gemini 1.5 Flash 002
34.03±1.41
25
Pixtral Large (November 2024)
33.89±0.69
25
32.69±1.40
29
Qwen2-VL-72B-Instruct
28.56±1.37
29
Claude 3 Opus
27.82±0.55
31
26.55±0.35
31
Nova Pro
26.27±0.61
31
Pixtral 12B (September 2024)
25.97±0.74
31
Nova Lite
25.50±0.77
33
Llama 3.2 90B Vision Instruct
24.61±0.80
36
Llama 3.2 11B Vision-Instruct
20.47±0.15
37
Phi 3.5 Vision-Instruct
15.18±0.81