The rollout, which expands what Veo 3 can do within the Gemini app, is not available to all users, only those on paid Ultra and Pro plans can access it, and even then, they’re limited to three creations per day.

The update builds on what Google had already introduced in its standalone tool, Flow, earlier in May. While Flow offered similar functionality, embedding this into the Gemini is a smart move to consolidate its generative tools under one roof, while tightening control over access.

You start with an image. You upload it, choose “Videos” from the menu, and type a prompt describing what you want to see and hear. The system then generates an animated 8-second video in 720p, complete with music, background noise, or even speech, depending on the prompt. 

Once done, you can download or share it. But then, you can’t bypass the subscription wall or exceed the three-video daily limit. There’s no rollover or extensions.

According to Google, over 40 million videos have already been generated using Veo 3 and Flow combined, since the system launched just seven weeks ago.

That level of adoption, even under a paywall, shows high demand from creators and marketers, though it leaves us wondering about accessibility and reach, especially in lower-income regions.

Every video comes stamped with two identifiers: a visible “Veo” watermark, and a hidden SynthID, a digital signature embedded in the content that allows it to be traced as AI-generated. Google says this dual-tagging approach helps to prevent misuse and maintain transparency.

The company added, “We’re committed to safe and responsible deployment. Our systems are tested through red teaming, and we enforce moderation protocols to avoid sensitive or deceptive outputs.”

What’s interesting is that Google has kept Veo 3 off-limits to free-tier users. While trial periods may occasionally be offered, the core model remains paywalled. Subscriptions for Veo 3 begin at $19.99 per month for Pro access, and scale up to $249.99 monthly for the Ultra plan. 

That high cost may make sense for enterprise users, but it’s a barrier for casual creators and educators hoping to explore visual storytelling.

The new image-to-video tool is now live in over 150 countries, including India, Indonesia, and most of Europe, pointing to a global strategy that prioritises high-population, mobile-first markets. 

Nonetheless, Google’s careful rollout implies it’s less interested in mass availability and more focused on tightly managing use cases as the technology matures.

With this rollout, subscribers in regions including Europe, India, and Indonesia can now create short videos using just text prompts. 

However, users are limited to three videos per day. After that, they’re pushed back to Veo 2, an older version with fewer features and no native audio.

This shows Google’s continued focus on product differentiation between its AI subscription tiers. Initially reserved for AI Ultra subscribers at $249.99/month, Veo 3 was quietly made available to AI Pro users (who pay $19.99/month) last month, albeit with access to the slightly slower “Veo 3 Fast” version.

Compared to Veo 2, the upgrades in Veo 3 are commendable. The tool now generates videos with built-in audio, including dialogue, ambient sounds, and even nature effects, giving creators more realistic, complete video outputs without needing post-production.

Josh Woodward, a lead product executive at Google, confirmed on Wednesday: “India, Indonesia, all of Europe, and more are starting to get access to create videos right now.”

Beyond audio, the model also introduces smoother motion, improved character consistency across frames, and more accurate lighting and physics, raising the bar for realism in short video generation. 

These enhancements address long-standing complaints about previous models’ choppy animations and visual glitches.

Google also claims users will face fewer interruptions when creating content, citing reduced content blocking. However, no specifics were given about how moderation thresholds have changed, a likely sensitive area as synthetic video tools come under checks globally.

Importantly, every video generated with Veo 3 is tagged with SynthID watermarks, both visible and invisible, to maintain traceability and discourage misuse. This is beyond protecting the tool’s integrity, but defending Google’s reputation as it wades deeper into generative content.

A later update will allow Google users to upload static images and animate them using Veo 3’s engine, an image-to-video feature already offered by some competitors like Runway and Pika. Unlike OpenAI’s Sora, which remains in limited preview, Google Veo 3 is publicly accessible to all paying users.

Veo 3’s most captivating use cases range from marketing visuals to concept design. According to Google, it’s aimed at creators “who want to break through creative blocks and visualise ideas in a flash.”

That may be true, but for now, they’ll have to visualise in under eight seconds.