Google DeepMind has unveiled Veo 3.1, a significant upgrade to its generative video model that emphasizes creative control and production-grade quality. The centerpiece of this update is "Ingredients to Video," a feature that transforms reference images into dynamic clips while maintaining striking consistency. By allowing users to provide "ingredient" images as a baseline, the model preserves the specific look of characters and environments across different prompts, addressing the persistent challenge of maintaining the same visual identity over time.

For creators targeting mobile platforms, Veo 3.1 now supports native vertical outputs in a 9:16 aspect ratio, eliminating the need for awkward cropping that typically degrades visual quality. This mobile-first approach is paired with advanced upscaling capabilities, enabling the generation of high-fidelity 1080p and 4K footage. Such improvements bridge the gap between casual experimentation and professional filmmaking, providing the sharp textures and clarity required for high-end productions and large-screen displays.

Integration across the Google ecosystem ensures these tools are accessible to everyone from YouTube Shorts creators to enterprise developers via the Gemini API. To maintain transparency, all generated videos include SynthID, an imperceptible digital watermark used to identify AI-generated content. This update signals a shift toward modular creativity, where AI acts as a sophisticated tool for precise visual storytelling rather than a random generator.

Google DeepMind has unveiled Veo 3.1, a significant upgrade to its generative video model that emphasizes creative control and production-grade quality. The centerpiece of this update is "Ingredients to Video," a feature that transforms reference images into dynamic clips while maintaining striking consistency. By allowing users to provide "ingredient" images as a baseline, the model preserves the specific look of characters and environments across different prompts, addressing the persistent challenge of maintaining the same visual identity over time.

For creators targeting mobile platforms, Veo 3.1 now supports native vertical outputs in a 9:16 aspect ratio, eliminating the need for awkward cropping that typically degrades visual quality. This mobile-first approach is paired with advanced upscaling capabilities, enabling the generation of high-fidelity 1080p and 4K footage. Such improvements bridge the gap between casual experimentation and professional filmmaking, providing the sharp textures and clarity required for high-end productions and large-screen displays.

Integration across the Google ecosystem ensures these tools are accessible to everyone from YouTube Shorts creators to enterprise developers via the Gemini API. To maintain transparency, all generated videos include SynthID, an imperceptible digital watermark used to identify AI-generated content. This update signals a shift toward modular creativity, where AI acts as a sophisticated tool for precise visual storytelling rather than a random generator.