Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Web Image Translation on Multilingual Web Image Translation TH-EN
Loading...
43.2
BLEU
GPT4.1†
0.768
11.784
22.8
33.816
May 23, 2026
BLEU
COMET
Updated 8d ago
Evaluation Results
Method
Method
Links
BLEU
COMET
GPT4.1†
Tuning Strategy=Zero-Shot
2026.05
43.2
96.2
Gemini2.5 Pro†
Tuning Strategy=Zero-Shot
2026.05
42.1
91.1
VaaWIT on Qwen3 (8B)
Backbone=Qwen3 (8B), T...
2026.05
39.8
87.5
Full Fine-Tuning
Backbone=Qwen3-VL (8B)...
2026.05
36.8
84.4
VaaWIT on LLaMA3.1 (8B)
Backbone=LLaMA3.1 (8B)...
2026.05
33.5
76.4
PP-OCR_Microsoft API
Type=Cascaded Model
2026.05
23.5
76.5
EasyOCR_Google API
Type=Cascaded Model
2026.05
20.3
69.8
Chain-of-Thought
Backbone=Qwen3-VL (8B)...
2026.05
18.6
71
Qwen3-VL (32B)
Tuning Strategy=Zero-S...
2026.05
14.3
69.9
Qwen3-VL (8B)
Tuning Strategy=Zero-S...
2026.05
12.2
64.8
LoRA
Backbone=Qwen3-VL (8B)...
2026.05
10.5
62.4
LLaMA3.2 (90B)
Tuning Strategy=Zero-S...
2026.05
5.5
48
LLaMA3.2 (11B)
Tuning Strategy=Zero-S...
2026.05
3.1
47.9
LLaVA-OV (7B)
Tuning Strategy=Zero-S...
2026.05
2.4
44.2
Feedback
Search any
task
Search any
task