Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Multimodal Conversation on LLaVA Bench

93.1LLaVA Bench Score

GPT-4V (2023.11.06)

9.69231.3465374.654Aug 3, 2024Nov 19, 2024Mar 7, 2025Jun 23, 2025Oct 9, 2025Jan 25, 2026May 14, 2026
Updated 15d ago

Evaluation Results

MethodLinks
93.1---
2024.08
86.7---
2026.05
85.9---
2026.05
83.7---
2024.08
83---
2024.08
82---
2024.08
81.8---
2026.05
80.8---
2026.05
80.2---
2024.08
80.1---
2024.08
79.9---
2026.05
78.2---
2024.08
77.8---
2026.05
75.3---
2026.05
74.8---
2026.05
74.8---
2026.05
74.7---
73.9---
2024.08
73.9---
2026.05
73.8---
2026.05
73.4---
2026.05
72---
2024.08
71---
2024.08
69.2---
2024.08
69.2---
2026.05
67.8---
2026.05
67.8---
2026.05
67.8---
2024.08
67.7---
2026.05
67.7---
2026.05
66.5---
2026.05
66.3---
2026.05
66---
2026.05
65.2---
2026.05
65.2---
2026.05
64.3---
64.2---
2024.08
62.3---
2024.08
61.2---
2026.05
55.5---
2024.08
51.9---
2024.08
51.3---
2024.08
51.1---
2024.08
49.1---
2026.05
22.3---
2026.05
12.9---
2023.12
-85.474.396.3
2023.12
-89.379.797.7
2023.12
-61.747.355
2023.12
-93.874.3111.4
-83.267.690.6
-81.977.192.3
2023.12
-81.675.595.2
2023.12
-93.175.391.6
2023.12
-96102.5106.7