Prompt Battle / Image-to-Video Battle
Image-to-Video Battle
● Results in productionImage-to-video · Same starting image · Category: Image-to-Video Tools
The starting image + motion prompt — identical for every tool
“Starting image: a still photo of a red vintage scooter parked outside a small Lisbon café at 8am, warm side light, a cat asleep on the seat. Motion prompt: the camera pushes in slowly; the cat wakes, stretches and jumps down; the café awning flutters in a light breeze.”
This wording is locked and was published before any generation started. Every contender gets exactly this input, on an entry paid plan, with settings as close to default as the tool allows.
The contenders
Runway
Entry plan from $12/mo · credits billing (from our Cost Calculator data)
Kling AI
Entry plan from $9/mo · credits billing (from our Cost Calculator data)
Leonardo AI
Paid plans available; free tier with daily tokens
How it will be judged
| Criterion | What we look at |
|---|---|
| Image fidelity | Does the moving clip still look like the starting photo — same scooter, same café, same cat? |
| Motion naturalness | The wake-up, stretch and jump: believable animal movement, or rubber physics? |
| Subject consistency | Do the scooter, cat and background stay themselves for the whole clip, or morph mid-shot? |
| Cost per usable clip | What the generation cost on a real entry plan, divided by the clips that were actually usable. |
Status: results in production
The outputs are being generated on real accounts; first results drop soon. There are no outputs, scores or winners to show yet — and there won't be until the real generations are finished and judged against the criteria above. When results land, this page will show every output unedited, what each run cost, and the usable-seconds count for each contender.
Email me when the results drop: