Google’s Veo 3: AI Video Generation Arrives

Creating ultra-realistic videos via AI has arrived in Portugal with **Google’s** Veo 3. The tool, unveiled at **Google** I/O, is accessible through the Gemini chatbot and offers powerful video generation capabilities.

Veo 3: From Text to Moving Images

Veo 3 constructs video files using text prompts or static images. This new iteration produces scenes with ambient sound, realistic dialogue, and varied visual styles, marking a jump over previous systems.

The AI model analyzes textual descriptions, using the instructions to generate realistic and creatively controlled video sequences. According to a 2024 report by Exploding Topics, the AI video market is projected to reach $23.58 billion by 2030 (Exploding Topics 2024).

How to Access Veo 3

The video-generating AI is part of Google AI Pro, a subscription service with advanced AI tools. Subscribers gain access to Gemini for assistance with documents and emails, analysis of large texts, and the ability to create videos via Veo 3.

A subscription costs approximately 22 euros monthly, with a one-month free trial. You can subscribe through various methods using a Google account, including directly via the Gemini chatbot at gemini.google.com.

Tips for Optimal Video Generation

Veo 3 allows users to generate videos up to eight seconds long; credits are limited, even with a Google AI Pro subscription. The more descriptive and detailed a prompt, the better the results.

Descriptive prompts minimize AI “hallucinations” and unwanted creative decisions. For example, when creating a character, detail their appearance, dialogue, and voice, in addition to camera movements and environmental details.

Example Prompt

Here is an example of a prompt used to generate a video:

An elderly fisherman, with wrinkled face and contemplative look, is sitting in a small aged wooden bench, on the quay of a quiet port at the sunrise. Wears typical fisherman clothes: spent flannel shirt, thick wool vest, blue waterproof pants and rubber boots. It uses a dark wool hat that partially covers gray hair. Hold a mug of nail polish with both shredded hands and drinking hot tea slowly, with the steam raising gently in the cold air of the morning.

The camera slowly approaches the fisherman’s face as he smiles slightly and, in a calm and deep voice, says:
“The sea does not teach with words, but with silence, patience… and respect.”

The sea is calm in the background, with golden reflexes of the rising sun dancing on the soft waves. Some fishing boats shake lightly anchored. The atmosphere is melancholy but serene, with a smooth color palette: blue, gray and hot golden tones. The ambient sound includes distant seagulls and the light waves on the pier.

With structured prompts, Veo 3 can produce videos of surprising visual quality and narrative complexity, formerly the realm of production teams and equipment. One limitation is that the AI still struggles with Portuguese pronunciation.