Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Slide Editing on Talk-to-your-slides Instruction Set Overall
Loading...
96.57
Success Rate (%)
TALK-TO-YOUR-SLIDES
72.9308
79.0679
85.205
91.3421
May 16, 2025
Success Rate (%)
Instruction Following Score
Text Quality Score
Image Quality Score
Layout Quality Score
Color Quality Score
Execution Time (s)
Average Input Scale (k)
Average Output Scale (k)
Cost (x10^-3)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Success Rate (%)
Instruction Following Score
Text Quality Score
Image Quality Score
Layout Quality Score
Color Quality Score
Execution Time (s)
Average Input Scale (k)
Average Output Scale (k)
Cost (x10^-3)
TALK-TO-YOUR-SLIDES
Backbone=GPT-4.1-mini
2025.05
96.57
2.13
2.46
2.26
2.71
2.68
78.37
3.6
1.89
3.8
Direct code gen.
Backbone=GPT-4.1-mini
2025.05
76.25
0.53
0.81
1.07
1.23
1.15
18.19
1.06
0.48
1.2
UI Agent
Backbone=GPT-4.1-mini
2025.05
73.84
1.68
1.94
1.55
2.16
2.04
121.08
98.22
2.3
15.4
Feedback
Search any
task
Search any
task