Google has brought Veo 3.1’s Ingredients to Video feature directly into YouTube Shorts and the YouTube Create app, putting its latest video generation model in front of the short-form creators most likely to use it. The update, detailed on Google’s official blog, also extends portrait-mode Veo generation to the Gemini app.

Ingredients to Video lets creators supply up to three reference images to steer a generation, locking in a character’s face, wardrobe colors or an environment’s lighting direction across consecutive shots. That character consistency has been one of the hardest problems in AI video, and it is the difference between a one-off novelty clip and a repeatable series with a recognizable cast. Combined with native 9:16 vertical output, creators can now produce full-screen vertical storytelling without cropping or reframing in an editor afterward.

The model itself focuses on depth, realism and creative control. Google points to sharper textures, better lighting, smoother motion and improved lip-sync precision for dialogue and voiceovers. A multi-prompt feature lets creators plan storyboards or chain several ideas into one continuous sequence, and scene extension can carry a narrative to a minute or more while maintaining continuity. State-of-the-art upscaling pushes output to 1080p and 4K for creators who need higher-resolution masters.

For professional workflows, the same updates are rolling out to Flow, the Gemini API, Vertex AI and Google Vids. Every video generated through Google’s tools carries an imperceptible SynthID digital watermark, and the Gemini app’s verification tool now accepts video uploads, so anyone can check whether a clip was made with Google AI.

The practical takeaway for Shorts creators is to treat Ingredients to Video as a pre-production tool: build a small library of reference images for your recurring characters, locations and visual motifs, then generate variations against that locked cast instead of rolling the dice on every prompt. Consistency compounds, and the creators who systematize it will look markedly more professional than those generating each clip from scratch.