Image Captioning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
78.6GPT-4 Consistency (Clean)
6
Mar 31, 2026
41.67Claude Score
6
Mar 31, 2026
63.7Score
6
Mar 6, 2026
70CIDEr
6
Feb 26, 2026
0.918BLEU-1
6
Feb 26, 2026
78.6Clean Score
6
Feb 26, 2026
78.3ASR (Average Sentence Rate)
6
Feb 26, 2026
38.4BLEU@4
6
Feb 26, 2026
15.7CIDEr
6
Feb 26, 2026
130.7CIDEr
6
Feb 26, 2026
127.9CIDEr
6
Feb 26, 2026
19.15GPT-Score
5
May 19, 2026
53CS Score
5
Apr 28, 2026
51.06TTS
5
Apr 3, 2026
77.25Score
5
Mar 20, 2026
63.64Unified Score
5
Mar 20, 2026
64.4CapArena Win Rate (vs Base Model)
5
Mar 11, 2026
28.59BLEU
5
Mar 6, 2026
61.4D1
5
Mar 6, 2026
39.6BLEU-4
5
Feb 26, 2026
75.55BS-P
5
Feb 26, 2026
0.394BLEU-4
5
Feb 26, 2026
133.3CIDEr
5
Feb 26, 2026
139.1CIDEr
5
Feb 26, 2026
34.9BLEU Score
5
Feb 26, 2026