The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
I used the same image-generation prompt across Gemini, ChatGPT, Grok, and Snapchat AI to compare how each platform interprets identical instructions. It was interesting to see how every model produced a unique result in terms of composition, realism, lighting, and overall style. The exact prompt used is included below for reference.
Prompt: The woman/Man in reference wearing baggy clothes looking at the camera from in between a motion blurred urban croud at shibuya crossing the woman is in his early 20s, there should be open pores, slight pigmentation and micro texture, the camera is placed at a 60 degree angle above the man and all the croud exept the man is moving motion blurred --chaos 5 --ar 3:4 --raw --stylize 250 --hd
#snapchat
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Trending
retrofuturism
Share your favorite designs from the past exploring the distant future. Or your own take on the aesthetic 🚀
halftone
Dots. Texture. Grain. Let's see your designs using the halftone technique.
Midjourney
Creatives are showing off what they're making with Midjourney.
colorgrade
Video editors sound off: How do you enhance the mood of your video in color grading?
Related posts
Is AI finally good enough to shoot a product demo on its own, or human touch is still needed?
I generated a 20s product demo with Seedance 2.5 in a single take providing it with one product image reference and an example video transcript (posted in the comments).
Do you disclose to a client how much AI is involved, or just let the result speak for itself?
The future is here
i made a 90-sec Hollywood-style motorcycle chase short, completely 100% AI-generated.
Two riders, five pursuers, one packed highway. No crew, no stunts, just
Seedance 2.5 and a very detailed prompt "bible" to keep 5 bikes chasing in
the right direction.
The hardest part wasn't the visuals, it was physics. Seedance 2.5 kept putting the chasers IN FRONT of the getaway bike instead of behind it. Then i had to engineer a whole camera system + spatial rules to fix it. Then redesigned the ending after gunfire got flagged by moderation.
what i have finally had peace with is; AI filmmaking is 90% prompt engineering.
also, check the comment section below to see my process video of how i made this.
The LEAP - my submission for the Seedance 2.5 Capcut Challenge on Contra as part of the CapCut Creative Partners program
In the future ‘trying out’ stuff in real life might start in VR. It might as well end up there. This is a story of desire for adventure and exploration, yet bounded in room. Real activities turned into digital entertainment. ‘Participating’ in different sports might be just a dropdown menu to select in a new game. Some might say that the is a leap is forward, some that is a leap backwards. That’s up to you to decide.
I’ve always wanted to do this exact story and Seedance 2.5 made it possible. 7 locked-in visual sheets to direct the exact outfits for each scene and the room environment for the end shot. Seedance 2.5 did a brilliant job at nailing all the reference looks!
Workflow:
1. I created each reference sheet to match the outfits I wanted for the sports I wanted, using Nano Banana Pro
2. Then I worked on the concept then into the prompt.
3. Did initial tests and I noticed where the prompt falls and wanted to change the ending (with the current one)
4. Did some prompt itterations until I finally landed on the ONE SHOT
Loom project walkthrough
https://www.loom.com/share/c9d2fbf4b449406eb6e37d5a1a35be13
prompt here:
https://docs.google.com/document/d/1L58QJHOg3raNMU5Oy6LrPropcUuvgANMUC7QjofuH_8/edit?usp=sharing
Adhd people choosing new hobby 😂