Mark KrynskyFounder, LIfeScribe.us · Digital Legacy Advocate. Building personal data projects using AI, sharing tools and learnings

Kling 1.5 vs Hailuo Image to Video Model Comparison

Name: Kling 1.5 vs Hailuo Image to Video Model Comparison
Uploaded: 2024-12-31T17:33:26+00:00
Duration: 26 s
Channel: Mark Krynsky
Description: A side-by-side look at Kling 1.5 and Hailuo's image-to-video models. One generates a person jumping impossibly high while a background figure levitates…

31 Dec 2024 · 0 min

Watch on YouTube

0:26

A side-by-side look at Kling 1.5 and Hailuo's image-to-video models. One generates a person jumping impossibly high while a background figure levitates and dances, with bodies contorting unnaturally. It technically follows the prompt but looks completely wrong.

So this is totally unrealistic with me jumping vertically and the guy in the background levitating. The guy in the background is doing some crazy dance and I'm just going into contortions. This actually resembles the second jump and the people in the background look very unrealistic. This adheres to the prompt well with me landing in the sand, but also looks very unrealistic. I included this just to show what happens when things go horribly wrong.