I finally found out what VEO 3 is useful for... ACTION FIGURE FIGHTS! ๐ฅ
After months of testing VEO 3 for serious work, its real strength turned out to be something much sillier: making toy action figures fight like they're in a movie.
Months of testing Google's VEO 3 for polished, "serious" video projects led to an accidental discovery: the model's real strength shows up when you point it at toy action figures instead.
An accidental find
The setup was simple. Two action figures, Chun-Li and a martial arts opponent, staged as if mid-fight in the street. What came back from VEO 3 wasn't a stiff, static toy photo brought to life. It was motion blur on every kick and punch, a shallow depth of field that isolates the fighters from the background, and lighting that makes plastic figures read as if they're lit for a real movie set. The camera tracks the action smoothly, with the kind of bokeh in the background usually associated with professional cinematography, not toy photography.
Why VEO 3 handles this so well
The qualities that make this work are the same ones that are hard to fake with practical toy photography or basic stop motion: convincing motion blur that sells speed and impact, cinematic depth of field that pulls focus onto the fighters, lighting that gives static plastic a sense of weight and life, and camera movement that tracks the fight rather than sitting locked off. Put together, those elements turn a simple toy setup into something that looks like a genuine fight scene rather than a photo of two figures posed on a table.
From childhood playtime to Hollywood-style content
There's something genuinely funny about spending months chasing "serious" use cases for an AI video model only to find that its most convincing output comes from restaging childhood action figure battles. But the result speaks for itself: the level of detail in these sequences, from signature character moves to the physics of a strike landing, is well beyond what a static toy shot could deliver. It's a reminder that sometimes a tool's best use case isn't the one you set out to find.
Key takeaways
- VEO 3's motion blur and depth of field are strong enough to make static action figures look like they're in a real fight scene.
- Lighting realism is a major factor in selling plastic figures as "alive" on screen.
- Smooth camera tracking adds a cinematic quality that's hard to achieve with basic toy photography.
- The most useful application of a new AI tool isn't always the one it was originally tested for.
Try it yourself
If you have action figures lying around, this is a low-effort way to test what an AI video model can actually do with motion and lighting. Pick two figures, stage a simple pose, and see how the model handles the transition into motion. It's a fast, cheap way to learn the tool's strengths before applying it to more involved projects.
More videos
2:08Bridging the Pixel Gap in Browser Automation.
2:23How One Narrow Safety Rule Can Make an AI Less Safe Everywhere Else.
2:04Why splitting a chunk from its document makes it retrieve for the wrong question
4:20Three You Can Take Back. One You Can't.
2:21Why a 50-turn agent pays for the same screenshot 35 times unless it caches the pixels
1:53