Overview
Google announces a major update for Flow, its AI-assisted video creation tool, with the introduction of the Veo 3.1 model. This version improves editing capabilities, adds audio to more features and gives the user greater narrative and visual control.
Since its launch five months ago, Flow has generated more than 275 million videos with the help of Veo. With this new version, Google responds to feedback from users who asked for more creative-control tools and better audio integration, so the resulting videos feel more complete and expressive.
What Veo 3.1 brings: improvements and new features
1. Richer audio across more features
One of the most significant changes is that audio now extends to all existing Flow features. The features where audio is integrated include:
- Ingredients to Video: lets you build a scene using multiple reference images. With Veo 3.1, the resulting video will carry audio consistent with the scene.
- Frames to Video: by providing a start image and an end image, Flow generates a smooth transition between them; now also with audio that accompanies the visual journey.
- Extend: to lengthen an existing shot; with Veo 3.1, that added segment will also include believable sound.
2. More precise narrative control and in-scene editing
Users get finer control to adjust and modify scenes directly within Flow:
- Insert: you can now introduce new elements (an object, character or detail) at any point in the scene. Flow integrates them naturally: it calculates shadows, lighting and visual consistency.
- Remove (coming soon): the ability to remove unwanted objects or characters. Veo will reconstruct the background so it looks like the element was never there.
3. Better prompt adherence and visual realism
Veo 3.1 improves the fidelity between the user's prompt and the visual result, bringing the generated video closer to what the user requested. It also reinforces audiovisual realism, capturing textures, light nuances, atmosphere and details that make the resulting scene look more believable.
Availability and use
The new capabilities are available within Flow from the day of the announcement. Veo 3.1 is also available through the Gemini API, Vertex AI and the Gemini app for developers and companies that want to integrate it into their own platforms.
Significance and practical value
These improvements empower creators to experiment, adjust and refine their video scenes without feeling tied to the first result. Having integrated audio, plus fine visual adjustments, lets an AI-generated video get closer to a complete multimedia production.
By offering Veo 3.1 via APIs and cloud services, Google makes it easier for other apps or services to integrate AI video generation. This opens up possibilities for marketing, education, content creation, entertainment tools and more.
Challenges and limitations
- Some features (like "Remove object") remain promises for future versions.
- In scenes with many elements or complex motion, the model may have limitations in maintaining full visual and acoustic coherence.
- Questions arise about authorship, licensing, the use of data as reference and attribution.
This announcement reinforces the trend that generative models not only produce content, but also offer automatic editing and refinement tools. Veo 3.1 represents a step forward toward more complete and interactive systems, where the user retains creative control even in generative environments.
