This is an AI lipsync generation comparison for
image-to-lip-synced-video 
prompt:
Yes, my shirt says “Punch Me." No, you won’t win. I dodge like a jazzy ninja on espresso. Tested by www.AIcreators.tools...
Log in to see full prompt.
Tested: May 25, 2025
Generated form image+text. Voice was generated for this avatar, selected from 6 variants.
Tested: May 25, 2025
Master mode, used prebuilt voice: Callum.
Tested: May 28, 2025
HunyuanVideo-Avatar's prebuilt voice: Domineering Zihao. Shorter text due to demo website's constraints.
Tested: June 24, 2025
Text read by inbuilt voice - Callum
Tested: October 4, 2025
It does render pretty quickly. Seems to like cutting out the last syllable. My test was longer in 1st attempt and got cut mid-word, so I shortened it. Still finishes abruptly.
Tested: May 3, 2026
Quite impressive for the speed of generation it gave especially
Using platform's prebuilt voices to generate speech from text and animate the image of a dude wearing a 'punch me' shirt.
Check out the results from HeyGen vs Dreamina AI (OmniHuman 1) vs HunyuanVideo Avatar vs Hedra (Hedra Character 3) vs Fal AI (Ovi) vs Replicate (P-Video-Avatar) for similar or identical prompts side-by-side.
