Get the verdict
Same-prompt test 7 · video · October 2026

Skateboard kickflip down stairs: which AI does it best?

Runway Gen-4.5 wins with 6.8/10. Behind it: Veo 3.1 Fast (6.0), PixVerse V6 (6.0), Kling 3.0 (5.5). One 4-second clip per model, unedited, with whatever audio the model made.

The promptSkateboarder doing a kickflip down stairs, tracking shot.
#ModelOveralladherencemotionqualityartifactsCostRender
1Runway Gen-4.56.87686$0.4873s
2Veo 3.1 Fast6.06576$0.6037s
3PixVerse V66.06576$0.2138s
4Kling 3.05.54576$0.52240s

Best per category: adherence Runway (7) · motion Runway (6) · quality Runway (8) · artifacts all tied (6)

Every output, ranked

#1 Runway Gen-4.5

Runway Gen-4.5 6.8
4s · ~48 cr ($0.48) · 73s render

adherence 7motion 6quality 8artifacts 6
Board flips, landing on flat ground; stairs barely part of the trick.

#2 Veo 3.1 Fast

Veo 3.1 Fast 6.0
4s · audio · ~60 cr ($0.60) · 37s render

adherence 6motion 5quality 7artifacts 6
Board flips over stairs but skater stays at the top; physics off.

#3 PixVerse V6

PixVerse V6 6.0
4s · audio · 48 PixVerse cr ($0.21) · 38s render

adherence 6motion 5quality 7artifacts 6
Board flips off a ledge, but the skater never really goes down the stairs; landing physics off.

#4 Kling 3.0

Kling 3.0 5.5
4s · audio · ~52 cr ($0.52) · 240s render

adherence 4motion 5quality 7artifacts 6
Rides a ledge; no kickflip, no stair descent.

Try Runway ↗

⚖️
Outputs are shown exactly as the API returned them (4s, 720p, default settings). Scores are editorial judgments on a fixed rubric, given without knowing which model made each output. One run is a snapshot, not a guarantee. Methodology

FAQ

Which AI is best for "AI video of a skateboard kickflip"?

In our October 2026 test of this exact prompt, Runway Gen-4.5 wins with 6.8/10, ahead of Veo 3.1 Fast (6.0), PixVerse V6 (6.0), Kling 3.0 (5.5). Runway Gen-4.5's output: Board flips, landing on flat ground; stairs barely part of the trick.

What exact prompt was used?

“Skateboarder doing a kickflip down stairs, tracking shot.” Every model got the same prompt with default settings, 4 seconds at 720p, and one best-effort retry only if a generation failed.

How were the outputs scored?

Each output was scored without knowing which model made it, from 1 to 10 on adherence, motion, quality, artifacts. The overall score is the average of those.