This is an AI video generation comparison for
text-to-video 
prompt:
Raw footage of a 22-year-old influencer screaming from excitement while flying with a parachute in the sky to impress his followers. Loose closeup on his screaming face, fisheye lens, crisp and candid raw footage...
Log in to see full prompt.
Tested: January 23, 2025
Realistic, raw footage + dynamic motion + wind animation video, larger resolution at https://youtu.be/FmVqiVjPnKw
Tested: January 31, 2025
Qwen2.5-Plus model's text-to-video raw footage simulation example with custom aspect ratio.
Tested: February 13, 2025
Realistic selfie footage test with Mochi 1 model through Genmo AI playground
Tested: February 16, 2025
Vidu 1.5 text-to-video raw selfie footage video example
Tested: August 6, 2025
Veo added a line for him to say
Tested: August 7, 2025
Oddly enough, POV in the prompt seems to make it less jittery, without it AI is trying to show a selfie stick
Tested: August 7, 2025
Model seems to need a more detailed prompt to generate this properly.
Face and emotion. A full-on scream locked at max excitement mouth wide eyes wild pure shot of adrenaline you can almost hear it.
Camera style and lens pick. Shot loose and close with a fisheye twist that stretches the face just a little makes everything feel extra loud and alive sky and blur all around.
Motion and air effects. Hair cheeks and clothes all whip back from the rush you catch tiny blurs from the speed hitting hard.
Lighting and look. Bright real sunlight hitting strong shadows crisp lines barely any edits so it keeps that raw almost shaky real-world feel.
Scene and setup. Flying through open sky earth way down below slight camera jitter that makes you feel like you’re right there in the chaos.
Check out the results from Hunyuan Video (HunyuanVideo) vs Qwen Image & Video Generator vs Mochi (Genmo AI) vs Vidu AI vs Google DeepMind's Veo (Veo 3 Fast) vs Sora (Sora) vs Vidu AI (Vidu Q1) for similar or identical prompts side-by-side.
