Model leaderboard
Every model runs the same Glif tests. Scores come from 2,042 blind votes by 10 Glif team members, plus 1,279 AI judge scores, each counting as half a vote. Each kind of task has its own board.
Score out of 100: how often people would pick this model's output over an average model's on the same tests. 50 is average.
How we score
Every model runs the same Glif tests. Team members see two outputs side by side without the model names and pick the one that better fits the test, or call it a tie. An AI judge also scores every output from 0 to 10.
A Bradley-Terry fit turns all those head-to-head results into one rating per model. Ties count as half a win, and each AI judge score counts as half a vote.
The score is the predicted chance people prefer this model's output to the average model's on the same board, so 50 is average. A board is one kind of task, like text-to-image. Scores only compare models on the same board.
The 90% range comes from resampling the votes and refitting many times. Overlapping ranges are close to a tie. A model ranks once it has 20 votes' worth of ratings on a board. Until then it shows no score, only how close it is.
Bars show each score’s 90% range. Overlapping bars are close to a tie.
Show vote countsImage · Text-to-image
View image modes308 team votes on 19 tests. A model ranks once it has 20 votes’ worth of ratings here.
1
GPT Image 2.5Glif default
85 out of 100, 90% range 79 to 91
2
Gemini 3.1 FlashGlif default
68 out of 100, 90% range 60 to 76
3
Nano Banana Pro
65 out of 100, 90% range 55 to 75
4
Seedream 5.0 Pro
63 out of 100, 90% range 52 to 73
5
FLUX 3 Image
59 out of 100, 90% range 47 to 69
6
FLUX.2 Pro
57 out of 100, 90% range 45 to 68
7
Muse ImageGlif default
55 out of 100, 90% range 46 to 66
8
Ideogram V4.5Glif default
55 out of 100, 90% range 42 to 66
9
FLUX.2 Flex
34 out of 100, 90% range 23 to 45
10
Krea 3Glif default
31 out of 100, 90% range 20 to 45
11
Seedream V4.5
31 out of 100, 90% range 20 to 46
12
Seedream 5.0 Lite
25 out of 100, 90% range 15 to 34
13
Grok Imagine Image
23 out of 100, 90% range 16 to 30
14
Z-Image Generator
17 out of 100, 90% range 9 to 25
Not ranked yet (1)
Video · Image-to-video
View video modes384 team votes on 3 tests. A model ranks once it has 20 votes’ worth of ratings here.
1
Wan 3.0
71 out of 100, 90% range 60 to 83
2
Wan 3.0 Reference
69 out of 100, 90% range 56 to 81
3
Hailuo H3 Max (MiniMax)Glif default
67 out of 100, 90% range 52 to 80
4
PixVerse V6 Image-to-Video
67 out of 100, 90% range 54 to 79
5
Seedance 2.0Glif default
66 out of 100, 90% range 53 to 79
6
Seedance 2.5
65 out of 100, 90% range 49 to 77
7
Kling V3 Standard · Image to video
62 out of 100, 90% range 48 to 76
8
Hailuo H3 Max Turbo (MiniMax)
60 out of 100, 90% range 50 to 72
9
Seedance 2.0 Fast
60 out of 100, 90% range 46 to 76
10
P-Video
60 out of 100, 90% range 49 to 72
11
Kling O3 Pro · Image to videoGlif default
42 out of 100, 90% range 29 to 57
12
Kling O3 Pro · Reference to video
42 out of 100, 90% range 31 to 53
13
Grok Imagine Video 1.5Glif default
42 out of 100, 90% range 25 to 57
14
Seedance Pro 1.5
41 out of 100, 90% range 28 to 51
15
Seedance 2.0 Mini
41 out of 100, 90% range 30 to 53
16
FLUX.3
41 out of 100, 90% range 28 to 55
17
Gemini Omni Flash 1.1
37 out of 100, 90% range 24 to 52
18
Kling V3 Pro · Image to video
37 out of 100, 90% range 25 to 50
19
Kling V3 4K · Image to video
35 out of 100, 90% range 25 to 47
20
Grok Imagine Video
33 out of 100, 90% range 22 to 47
21
Kling 2.5 Pro · Image to video
28 out of 100, 90% range 18 to 39
22
Gemini Omni Flash 1.1 · Reference to video
23 out of 100, 90% range 13 to 33
Audio · Text-to-music
View audio modes8 team votes on 2 tests. A model ranks once it has 20 votes’ worth of ratings here.
No ranked models yet. A model ranks once it has 20 votes’ worth of ratings here.