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.

Hide vote counts

384 team votes on 3 tests. A model ranks once it has 20 votes’ worth of ratings here.

Score (50 is average)
Credits per run, log scale
Vertical lines show each model’s 90% score range. Horizontal lines show its estimated cost range. Highlighted are the Glif defaults: Hailuo H3 Max (MiniMax), Seedance 2.0, Kling O3 Pro · Image to video, Grok Imagine Video 1.5.
  1. 1

    Wan 3.0

    90% range 60–83
    Wins 21 / Losses 7 / Ties 0 · 28 team votes · 6 AI judge scores

    71 out of 100, 90% range 60 to 83

  2. 2

    Wan 3.0 Reference

    90% range 56–81
    Wins 19 / Losses 8 / Ties 0 · 27 team votes · 7 AI judge scores

    69 out of 100, 90% range 56 to 81

  3. 3

    Hailuo H3 Max (MiniMax)Glif default

    90% range 52–80
    Wins 22 / Losses 9 / Ties 1 · 32 team votes · 7 AI judge scores

    67 out of 100, 90% range 52 to 80

  4. 4

    PixVerse V6 Image-to-Video

    90% range 54–79
    Wins 21 / Losses 8 / Ties 1 · 30 team votes · 7 AI judge scores

    67 out of 100, 90% range 54 to 79

  5. 5

    Seedance 2.0Glif default

    90% range 53–79
    Wins 20 / Losses 9 / Ties 0 · 29 team votes · 5 AI judge scores

    66 out of 100, 90% range 53 to 79

  6. 6

    Seedance 2.5

    90% range 49–77
    Wins 20 / Losses 11 / Ties 0 · 31 team votes · 5 AI judge scores

    65 out of 100, 90% range 49 to 77

  7. 7

    Kling V3 Standard · Image to video

    90% range 48–76
    Wins 19 / Losses 10 / Ties 0 · 29 team votes · 7 AI judge scores

    62 out of 100, 90% range 48 to 76

  8. 8

    Hailuo H3 Max Turbo (MiniMax)

    90% range 50–72
    Wins 20 / Losses 12 / Ties 0 · 32 team votes · 7 AI judge scores

    60 out of 100, 90% range 50 to 72

  9. 9

    Seedance 2.0 Fast

    90% range 46–76
    Wins 18 / Losses 11 / Ties 1 · 30 team votes · 5 AI judge scores

    60 out of 100, 90% range 46 to 76

  10. 10

    P-Video

    90% range 49–72
    Wins 17 / Losses 10 / Ties 4 · 31 team votes · 7 AI judge scores

    60 out of 100, 90% range 49 to 72

  11. 11

    Kling O3 Pro · Image to videoGlif default

    90% range 29–57
    Wins 11 / Losses 15 / Ties 2 · 28 team votes · 7 AI judge scores

    42 out of 100, 90% range 29 to 57

  12. 12

    Kling O3 Pro · Reference to video

    90% range 31–53
    Wins 11 / Losses 17 / Ties 2 · 30 team votes · 7 AI judge scores

    42 out of 100, 90% range 31 to 53

  13. 13

    Grok Imagine Video 1.5Glif default

    90% range 25–57
    Wins 12 / Losses 16 / Ties 1 · 29 team votes · 8 AI judge scores

    42 out of 100, 90% range 25 to 57

  14. 14

    Seedance Pro 1.5

    90% range 28–51
    Wins 11 / Losses 16 / Ties 2 · 29 team votes · 7 AI judge scores

    41 out of 100, 90% range 28 to 51

  15. 15

    Seedance 2.0 Mini

    90% range 30–53
    Wins 12 / Losses 17 / Ties 2 · 31 team votes · 5 AI judge scores

    41 out of 100, 90% range 30 to 53

  16. 16

    FLUX.3

    90% range 28–55
    Wins 11 / Losses 18 / Ties 0 · 29 team votes · 8 AI judge scores

    41 out of 100, 90% range 28 to 55

  17. 17

    Gemini Omni Flash 1.1

    90% range 24–52
    Wins 11 / Losses 19 / Ties 0 · 30 team votes · 7 AI judge scores

    37 out of 100, 90% range 24 to 52

  18. 18

    Kling V3 Pro · Image to video

    90% range 25–50
    Wins 11 / Losses 17 / Ties 2 · 30 team votes · 7 AI judge scores

    37 out of 100, 90% range 25 to 50

  19. 19

    Kling V3 4K · Image to video

    90% range 25–47
    Wins 10 / Losses 19 / Ties 1 · 30 team votes · 7 AI judge scores

    35 out of 100, 90% range 25 to 47

  20. 20

    Grok Imagine Video

    90% range 22–47
    Wins 8 / Losses 18 / Ties 5 · 31 team votes · 7 AI judge scores

    33 out of 100, 90% range 22 to 47

  21. 21

    Kling 2.5 Pro · Image to video

    90% range 18–39
    Wins 8 / Losses 20 / Ties 4 · 32 team votes · 7 AI judge scores

    28 out of 100, 90% range 18 to 39

  22. 22

    Gemini Omni Flash 1.1 · Reference to video

    90% range 13–33
    Wins 6 / Losses 23 / Ties 2 · 31 team votes · 7 AI judge scores

    23 out of 100, 90% range 13 to 33

Not ranked yet (3)