Google rolls out its new Veo 3 video‑generation model globally

Veo 3

Prime Highlights

  • Google has rolled out its AI video-generation model, Veo 3, globally to Gemini users in 159+ countries.
  • It’s only available to Google AI Pro subscribers with a limit of three 8‑second videos per day.

Key Facts

  • Veo 3 allows text-prompted video generation up to 8 seconds with synchronized audio.
  • Model introduced on July 3, 2025, under Gemini to pay subscribers of AI Pro.
  • Google is going to add features like image-to-video generation in a future update.

Key Background

Veo is a generative text-to-video sequence pioneered by Google DeepMind. It was first rolled out at Google I/O 2024 as a high-definition video model, with Veo 2 following in December 2024 that could handle 4K resolution and steadier physics and framing. In May 2025, Google launched Veo 3, which represents a step up by incorporating fully synchronized audio—music, ambient noise, sound effects, and even dialogue—on the erstwhile silent video shorts. DeepMind CEO Demis Hassabis termed the development as the end of the “silent film era” of AI-generated video.

Veo 3’s core innovation is this seamless integration of audio and visuals, a feature unmatched by earlier models like OpenAI’s Sora. During its unveiling, Google also introduced Flow, a companion tool that stitches together multiple short clips into a narrative sequence, assisting creators in visualizing scenes through an eight‑second clip builder .

Since launch, Veo 3 has gone viral: users are producing AI-generated scenes from animations to dialogue-driven shorts. Elon Musk even called it “awesome,” highlighting its filmic quality. But there are fears about possible abuse: TIME magazine learned that Veo 3 can create nearly unrecognizable deepfakes of riots or election tampering, enhanced by the realism of its audio as well as lifelike video. Google has attempted to limit these threats via watermarking—both imperceptible SynthID and an overt marker—but experts say stricter protections are still needed amid growing disinformation concerns.

By and large, Veo 3’s worldwide release to more than 159 nations under Google AI Pro is a milestone moment in generative media—bringing with it potent creative powers but requiring immediate innovation of ethical guardrails to avoid abuse.