Image Courtesy: Google

October 15, 2025 - Google DeepMind has announced Veo 3.1, the latest version of its cutting-edge generative video model, along with a major update to Flow, its AI filmmaking platform that empowers creators to turn ideas into cinematic experiences.

The update brings richer audio, more creative control, and enhanced realism, giving filmmakers, developers, and digital artists new ways to produce sophisticated, lifelike videos entirely from text and image inputs.

“We’ve been inspired by the creativity Flow has sparked with over 275 million videos generated,” said Jess Gallegos, Senior Product Manager at Google DeepMind. “With Veo 3.1 and new Flow capabilities, creators now have more narrative control and higher fidelity than ever before.”

What’s New in Veo 3.1

Building on the foundation of Veo 3, the new Veo 3.1 model delivers major improvements in prompt adherence, audiovisual quality, and texture realism. It also introduces generated audio for the first time, allowing Flow users to craft immersive soundscapes that match their visual storytelling.

Key Improvements in Veo 3.1:

• Richer, more realistic textures that capture lifelike detail across lighting, skin, and motion.

• Enhanced prompt accuracy for more faithful adherence to creative direction.

• Integrated audio synthesis, generating background sound, ambiance, and motion-linked audio directly from prompts.

• Seamless image-to-video realism, ideal for storyboarding and cinematic sequences.

Developers can now access Veo 3.1 through the Gemini API, Vertex AI, and the Gemini app, extending its reach to enterprise, app developers, and AI researchers.

Flow Gets Smarter: Creative Control Meets Audio

The Flow platform—Google’s AI video creation environment—has received its largest upgrade yet, integrating Veo 3.1’s new model capabilities across its most popular features.

Audio Comes to Every Feature

Audio is now available across all Flow capabilities, giving creators the power to design full scenes with synchronized visuals and sound:

• Ingredients to Video — Combine multiple reference images to define the style, objects, and characters in your scene. Flow blends these “ingredients” into a cohesive, visually accurate video.

• Frames to Video — Create smooth transitions between a starting and ending image, perfect for cinematic sequences or product demos.

• Extend — Generate seamless longer shots, connecting previous clips to build continuous, minute-long scenes.

New Editing Tools: Insert, Remove, and Reimagine

Responding to user feedback, Google has also added editing capabilities directly within Flow, allowing creators to refine their videos without leaving the platform.

• Insert — Add new objects, characters, or details mid-scene. Flow’s improved light and shadow modeling ensures inserted elements appear naturally integrated.

• Remove (coming soon) — Delete unwanted objects or characters while Flow reconstructs the scene background intelligently, maintaining consistency.

These tools allow for iterative, non-destructive video editing—opening up possibilities for cinematic precision, prototyping, and creative experimentation.

“Great ideas often evolve mid-creation,” said Thomas Iljic, Director of Product Management at Google Labs. “Our goal is to make Flow a fully creative canvas—one where creators can edit, extend, and refine without limits.”

For Developers: Veo 3.1 in Gemini API and Vertex AI

Veo 3.1’s technology is now available for developers and enterprises:

• Gemini API v2 gives developers direct access to Veo’s generative video and audio capabilities for integration into apps, tools, and creative pipelines.

• Vertex AI v3 enables enterprise customers to deploy custom AI video generation solutions securely at scale.

These integrations allow developers to harness Veo 3.1’s audiovisual generation capabilities for use cases like AI-powered video editing, advertising, education, and content simulation.

Expanding AI Filmmaking for All

With Veo 3.1 and Flow’s new editing suite, Google DeepMind continues to lead in AI filmmaking innovation. The combination of text-to-video generation, on-the-fly editing, and synchronized audio marks a significant step toward end-to-end AI production pipelines—tools that give creators cinematic control once reserved for professional studios.

Developers and creators can start experimenting with the new capabilities today in Flow or via the Gemini API.

“We’re opening up new possibilities for richer, more powerful storytelling,” Gallegos added. “With every update, we’re pushing AI filmmaking closer to true creative collaboration.”

Availability

The Veo 3.1 model is also available via the Gemini API for developers, Vertex AI for enterprise customers, and the Gemini app. New capabilities are available in both Gemini API and Vertex AI.