Google is significantly expanding access to Veo 3, their cutting-edge AI video generator, making it available to more users in numerous countries. This highly anticipated rollout means more people can now tap into the power of AI to create stunning, lifelike videos with synchronized audio.
Veo 3 represents a significant leap forward in AI video generation, distinguishing itself with the groundbreaking ability to generate not just visuals, but also natural-sounding dialogue, background noise, and music, all from simple text prompts.
This integration of audio directly into the generation process creates a far more immersive and realistic output.
Text-to-Video with Native Audio
The days of silent AI-generated videos are behind us. Veo 3 allows you to transform your text prompts into rich, realistic videos that automatically include sound effects, ambient noises, and even character dialogue that’s perfectly lip-synced. This opens up a whole new world of immersive storytelling possibilities.
Cinematic Camera Movements
Take control of your virtual camera like never before. Veo 3 offers a range of sophisticated cinematic camera movements and angles, from dynamic zoom-ins to fluid tracking shots. This empowers users to create professional-looking videos without needing extensive filmmaking expertise.
Consistent Characters and Objects
Maintaining visual consistency across frames has often been a challenge in AI video generation. Veo 3 tackles this head-on, ensuring that characters, environments, and objects remain consistent throughout your generated clips, even if you provide reference images, leading to more cohesive narratives.
Prompt-Based Editing and Control
Don’t settle for the first result. Veo 3 allows you to fine-tune your generated videos by providing additional prompts to adjust the style, tone, and overall structure. You can even use your own body, face, and voice as references to animate characters or precisely remove unwanted objects.
Josh Woodward, VP of Google, Google Labs, GeminiApp shared further details on the expansion.
Gemini Pro subscribers will get a trial pack of Veo 3 on the web (mobile soon), while Gemini Ultra subscribers get the highest number of Veo 3 gens w/ refreshes
If you’re a filmaker looking to exploit AI video generation, then consider Flow (http://flow.google): Pro: 10 gens/month, Ultra: Now 125 gens/month (up from 83).
Google says they know there are some issues, but wanted to get this into the hands of consumers as soon as possible.
One of the big changes in Gemini’s VEO3 text-to-video model is its ability to generate audio, along with the video. Today, Google only officially supports English audio output.
Testing Gemini VEO3
I already pay for Google One, mainly to get 2TB of photo backups, so I get Gemini Pro/Advanced included as part of that monthly cost. Having seen this post, I was keen to try it out, so asked Grok (sorry Gemini) for a prompt to test out VEO3.. here’s what it gave me.
Prompt: Create a 10-second cinematic video with synchronized audio, set in a vibrant futuristic city at dusk. The scene follows a young woman with short, neon-blue hair, wearing a sleek silver jumpsuit, as she walks confidently through a bustling street filled with holographic billboards and flying drones. A lively electronic music track plays in the background, with subtle sound effects of humming drones and distant city chatter. Midway through, she stops to look at a holographic advertisement displaying a dancing robot, and she smiles slightly. The camera pans to a close-up of her face, capturing her expression, then zooms out to show the cityscape glowing with neon lights. Ensure the visuals are crisp, with realistic lighting and physics, and the audio includes clear dialogue where she says, ‘This city never sleeps.’ The aspect ratio should be 16:9, and avoid generating any unrealistic distortions or artifacts.
I first went hunting for VEO in the model selection dropdown in Gemini, but seen nothing, but after visiting Flow, I was able to select Veo 3 from the quality settings menu on the top-right of the prompt, not exactly intuitive, but we got there.
By default, it’ll create 2 variants of the video, with each generation using this preset, consuming 100 AI credits. Given Pro users get 1,000 for the month, you’ll need to think carefully before using them.
Top-up AI Credits are only available to Google AI Ultra subscribers (in the US) and for 5,000 you’ll pay US$1,000. This means credits are $10 per 1,000.
The resulting video is pretty impressive (see below), particularly the inclusion of voice as instructed by the prompt.
The initial video generation happened quickly, in just seconds, and downloading it in 720p is also fast, however exporting an upscaled version to 1080, takes multiple minutes.
Let us know your thoughts in the comments.
The resulting Veo3 video
“Veo 3 is a game-changer for creators, allowing them to bring their ideas to life with unprecedented ease and realism. The ability to generate synchronized audio alongside high-quality visuals truly sets it apart and lowers the barrier to professional-grade video production.”
Tom Hume, Director of Product Management, Google DeepMind.
Sign in at http://gemini.google or http://flow.google with a Pro or Ultra account to get started.
