I finally got to use Google’s Veo 2 to generate AI videos and it’s better than Runway, Kling, Hailuo and Sora.
🔷 It’s faithful to your prompts
What you type is what you get. I used my test prompts such as “frog with pink wig writing on a parchment”, “chrome toasters with flapping white bird wings”, “asian monk meditating in a pool of radiating rainbow water”, and the results are quite accurate. The video quality is only 720p but I’m sure 1080p isn’t far away.
🔷 How to access Veo 2:
You can use Veo 2 within the Gemini Advanced app but you have to sign up for the Google AI Premium plan (one month trial, then S$29 monthly) which is much more expensive than my current Google One $9 monthly plan.
I don’t know if I’ll use much of the AI features (Gemini Flash 2.5, NotebookLM Plus) but it’s nice to have 2TB of cloud storage instead of 200GB. I’ll probably revert to my old Google One plan after a month because I’m already paying for Claude, ChatGPT and Midjourney.
Veo 2 is also available in Adobe Firefly, but I’ve abandoned Firefly as it’s too expensive and not useful enough. You can also use it through Gen AI platforms like Poe, but I believe it’s about $4 USD per video.
Bottom line: Your best bet is to use the free one-month trial of Google AI Premium.
🔷 It’s capped at about 10 videos a day within Gemini Advanced
I thought I was going to get unlimited video generations, but after generating about 8-10 videos, Google asked me to come back tomorrow. I actually feel excited about doing so, because Veo 2 is fast (1-2 minutes per video) and it doesn’t create crappy output like the other Gen AI video apps.