Gemini Advanced users can now generate 8-second, 720p videos via Veo 2, Google's advanced video model. The tool translates text prompts into dynamic clips, with examples including glacial caverns and voxel-style ice cream melting. Whisk Animate expands this capability, letting subscribers animate static images into videos using both text and image inputs. Both features are exclusive to Google One AI Premium subscribers, with rollout starting today across web and mobile in supported languages.

Veo 2 emphasizes cinematic realism, leveraging improved physics and motion understanding for lifelike scenes. Users simply describe a concept, and Gemini refines the output based on detail richness. A monthly video creation limit applies, though specifics weren't disclosed. Sharing options include direct uploads to TikTok and YouTube Shorts, enabling quick distribution of generated content.

Whisk Animate, initially launched in December as an image-generation tool, now integrates Veo 2 for animation. This update allows users to transform static visuals—like a mouse reading under a mushroom light—into fluid video sequences. Safety protocols include SynthID watermarking to denote AI origin and red teaming to mitigate policy violations. The feature launched in over 60 countries, though accessibility details remain sparse.

The updates reflect Google's push into generative media, competing with tools like OpenAI's Sora. By embedding safety measures and enabling cross-platform sharing, Veo 2 positions itself as a versatile creative suite. Users can test the tools at gemini.google.com and labs.google/whisk, with feedback mechanisms via thumbs up/down buttons to refine outputs.