After 1,000+ hours of testing & experimenting.

My list of AI for videos, from best to worst:

1. Google Veo 3 - gemini.google.com

- Best model to follow your prompt instructions.

2. Kinetix - kinetix.tech

- The best model to control the camera in 3D

3. Kling 2.5: lnkd.in/eB5fjyhd.

- The most creative model for image-to-video.

4. Hailuo - hailuoai.video

- Good prompt adherence, physics, consistency.

5. Pixverse: lnkd.in/gyJscdqN

- Good attention to detail and gets physics right.

6. Midjourney - midjourney.com

- The best model to make image (not video).

7. Luma AI - lumalabs.ai

- They released 'Ray-3', but I couldn’t make it work.

8. Moonvalley - lnkd.in/g5QESH7W

- A new player, but still not convincing.

9. OpenAI Sora - sora.chatgpt.com

- The hype was real. But it did not live up to it.

I made a blog where I test all of them. Read the full breakdown here: ruben.substack.com/p/vi…