AI VFX and experiments

144 of 3,062 videos

450 AI VFX and experiments, posted in public on X and Threads by the 42 creators who made them. Open any video to see who made it and what they used.

AI transforms middle-aged man into young woman

Hirokazu Yokohara1.1MX

Instead of checking my phone first thing, I set up an agent to make a "morning newspaper". It prints while I sleep. I used grokbot @bot to pull from my calendar and email I made it available if you wanna try it 👇

Karen X. Cheng1MX

AI virtual try-on from selfie and garment photos

Hirokazu Yokohara779KX

AI-only Blender desk scene test

Hirokazu Yokohara723KX

up next fabric/organic material experiments with Astra stay tuned

Fable 5.1 v GPT- 6 Astra im v impressed

Kling 3 vs Seedance 2 Original image generated by Seedream 5 Prompt: Live-action cinematic sequence. Another young woman, wearing a luxurious black two-piece bathing suit, enters the scene from the left (inside the house) and closes the drapes as they float in the gentle breeze. The woman walks, a

Alex Patrascu438KX

Late-night Veo 3 street interviews will spawn viral clips so hard. We live in a world of prompts:

Alex Patrascu428KX

Veo 3 vs Kling 2.1 Master Let's put them head-to-head in a cinematic battle. One re-roll per prompt. The sound is from Veo 3, no dialogue. Let's dive in:

Alex Patrascu316KX

Worth it! Seedance 2.0 Prompt below:

Alex Patrascu268KX

AI-modeled fantasy witch tower in Blender

Hirokazu Yokohara223KX

Testing Veo 3- Looping prompt below 👇

AI-made Blender office scene test, daytime

Hirokazu Yokohara190KX

🎥👩‍🎨⚗ In studio Paranormal - We are developing pipeline for use of AI in movies and games. If you want to help us build these future techniques, let us know! We don't accept any betatesting now, but we are open for investment or development collab🤗

Jakub Javora172KX

We shipped a small update to Story Panels! Cinematic stories will now look even better, and it also now supports the triptych logic from the Character Renderer.

Nicolas Neubert151KX

Testing Veo 3- Transitions pushing the limits of Veo 3 and i think i may have found an edge case!! I'm having a hard time getting the woman to actually hit the ball without disappearing behind it... so so close though will share the prompt and JSON shortly

So... I guess it's a ReAl 3D MoDeL now. (Unfortunately, the pipeline now takes 2 clicks and ~5 minutes to create a 3D character.)

Nicolas Neubert125KX

Look famous using Seedance 2.0 Prompt in reply.

CHRIS FIRST110KX

SITUATION: Astra has been working autonomously for 12 hours - (without me asking it to 😅) And i'm not stopping it. WHAT I INITIALLY ASKED FOR: >Simulate building an 18-hole course >Find a real location >Use maps + actual topographic data > Fit the holes to the natural terrain WHAT HAPPENED: >Rese

Rory Flynn83.6KX

Opus 5.5 rendered this in HTML from (1) image 🤯 Things are getting weird. PROMPT: We're going to try a little test. Do you think you could render a recipe motion graphic animation using javascript or html (w/e you think will produce the best) to show the full recipe from start to finish (empty gla

Rory Flynn80.4KX

More Wan 2.2 tests on Here I wanted to compare it directly to Veo 3 and also see how tight into the eye I could get with complex zooms. Again I’m really surprised by what Wan 2.2 can do. The prompts are the same across models, so Wan seems to be able to do more with the text, at

CoffeeVectors61KX

Not a single weak character 😭 Made with Pollo Recast by @itsPolloAI

Ciri60.1KX

fun experiment creating my Ai stylist, Sally👚✨ i used Codex + @VivixLabs_HQ to bring her to life from a character image, then asked her to style me for an event very interesting to see how it's able to reliably communicate what outfits i should go with test it here ⬇️

Minimaxxing Midjourney. H3 lip-sync is 👌 Also...this new MJ Edit model is 🔥

Rory Flynn44.3KX