Gemini Omni is a video generator in search of a workflow

Google dropped Gemini Omni into the $20 Gemini Pro plan this week, giving Pixel owners a way to generate and edit video from text, images, audio, or existing clips. The early takes call it ‘Nano Banana for video,’ and the demo reel looks sharp. But after a few hours of hands-on from people who got access, the story is already shifting. A Reddit user burned through their entire 5-hour usage window in one go and said the output ‘isn’t any better than VEO 3.1.’ The 1,000 Flow Credits a month get you roughly 25 edits, and once they’re gone, you wait.

That credit scarcity matters more than the quality debate. I wrote about a similar pattern with coding tools in The adoption gap — a tool can be technically impressive without ever slotting into someone’s actual production chain. For video generation, the line between ‘incredible demo’ and ‘usable asset’ is still wide enough to drive a truck through. If a model can’t give me a clip I can drop into a client project without burning my monthly allowance on one test, it’s a toy until the quota changes.

The Reddit thread about the launch perfectly captures the split: some people see a huge step forward because it’s multimodal and chat-native, while others point out that the model’s real strength is manipulation, not net-new generation, and it’s still locked behind a gate that punishes exploration. That tension is the only thing I’m watching now — whether Google widens the gate fast enough for anyone to actually build a workflow around this, or leaves it as a Pixel novelty that looks great in a tweet.