Google unveils Veo 3.1 with cinematic realism and richer storytelling for AI video generation

Google Veo AI model

Google has announced Veo 3.1, the latest version of its AI video generation model, alongside major updates to its Flow filmmaking tool. The new release promises enhanced realism, richer audio, and more control over cinematic storytelling — a clear step toward professional-grade AI video creation for creators and developers alike.

Building on Veo 3, which debuted at Google I/O 2025, Veo 3.1 introduces improved texture rendering, lifelike visuals, and richer, context-aware audio. The model now better understands narrative elements such as character interaction, pacing, and cinematic framing, allowing users to create more coherent and emotionally engaging stories from text or image prompts.

The image-to-video feature has been upgraded to deliver sharper visuals and smoother transitions, aligning closely with user instructions. Veo 3.1 and its faster variant, Veo 3.1 Fast, are now available across the Gemini app, Gemini API, and Vertex AI, supporting both horizontal (16:9) and vertical (9:16) video formats for creators working on everything from social content to film prototypes.

New features, tools in Flow

Alongside the new model, Google’s Flow filmmaking tool is receiving significant upgrades to harness Veo 3.1’s capabilities. Audio generation now extends to features like Ingredients to Video, which blends multiple images or characters into one seamless scene; Scene Extension, which continues an existing clip; and Frames to Video, bridging two images into a fluid motion sequence.

New editing features in Flow allow users to insert elements such as characters or objects while Google automatically adjusts shadows and lighting for realism. Coming soon is object and character removal — a feature that reconstructs backgrounds intelligently for a clean, natural look.