
What you need to know
- Google is bringing Veo 3's Ingredients to Video feature to the Gemini app starting today.
- The feature first debuted in Flow, and allows users to guide video prompts using reference photos.
- The feature is available for Gemini users on the Google AI Plus, Pro, or Ultra subscription plans.
Enjoy our content? Make sure to set Android Central as a preferred source in Google Search, and find out why you should so that you can stay up-to-date on the latest news, reviews, features, and more.
AI video creation is all the rave, with companies using the technology to create advertisements and build social platforms. Google recently rolled out the Veo 3.1 video-generation model to Gemini app users with enhanced realism, richer audio, and more narrative control. Now, it's adding expanded support for photo-to-video features in the Gemini app, as reported by 9to5Google.
The feature is called "Ingredients to Video," and it was announced alongside the Veo 3.1 release on Oct. 15. With it, users can provide Veo 3.1 reference images that will "control the characters, objects, and style." Ingredients to Video debuted in Flow, which is Google's dedicated AI filmmaking tool, but is rolling out in the Gemini app starting today.
Uploading reference photos (they can be AI-generated with Imagen, too) using Ingredients to Video can affect the final aspects of Gemini's output:
- Character Consistency: Maintaining the exact appearance of a character across different scenes or shots.
- Style Transfer: Applying specific textures, lighting, or artistic styles from a reference image to the entire video.
- World-Building: Ensuring objects and scenes in the video match a user’s custom-built world.
In a related move, Google update the Gemini app to show precisely which video generation model is being used to create your content. It appears in the Gemini Tools menu, and in this case, will show Veo 3.1.
Veo 3 in Gemini still has major limitations

Ingredients to Video is designed to make video generation simpler and quicker by cutting down the length and detail required of text-based prompts. Still, Veo 3 video generation is more limited than competing options, like OpenAI's Sora. Gemini and Veo 3 will refuse to generate certain videos, while Sora has looser guardrails.
The simpler way to generate videos with Veo 3.1 in the Gemini app is rolling out now, and Google says it will reach all paid users next week. At least for now, the feature is exclusive to Google AI Plus, Google AI Pro, and Google AI Ultra subscribers.