#1
Supporters
0
Support rate
0%
Combined
63
Task-specific approval voting
Models developers would actually use_
Choose a task, then support any number of specific model tiers. Objective evidence leads small samples; community influence grows with independent participants.
Participants
0
Community weight
0%
Sample status
Objective-led
My supported models
0
Current Top 20
Task leaderboard
Support is recorded per concrete model tier. Support rates can add up to more than 100% because one person may support several models.
#2
Supporters
0
Support rate
0%
Combined
59
#3
Supporters
0
Support rate
0%
Combined
56
#4
GLM 5.2
Z.ai · Z.ai / GLM 5.2
Supporters
0
Support rate
0%
Combined
53
#5
Supporters
0
Support rate
0%
Combined
48
#6
Gemini 2.5 Pro
Google · Gemini 2.5 / Pro
Supporters
0
Support rate
0%
Combined
48
#7
Supporters
0
Support rate
0%
Combined
45
#8
Supporters
0
Support rate
0%
Combined
44
#9
Supporters
0
Support rate
0%
Combined
44
#10
Supporters
0
Support rate
0%
Combined
44
#11
Supporters
0
Support rate
0%
Combined
43
#12
Supporters
0
Support rate
0%
Combined
41
#13
Supporters
0
Support rate
0%
Combined
39
#14
Supporters
0
Support rate
0%
Combined
39
#15
Supporters
0
Support rate
0%
Combined
39
#16
Supporters
0
Support rate
0%
Combined
37
#17
Supporters
0
Support rate
0%
Combined
37
#18
Supporters
0
Support rate
0%
Combined
35
#19
Supporters
0
Support rate
0%
Combined
35
#20
Supporters
0
Support rate
0%
Combined
35
One anonymous device can support each model once per task and can withdraw at any time. Salted IP rate limits and anomaly events protect the results.