Google Flow transforms AI creativity with conversational agents and mobile apps at I/O 2026

Google Flow evolves into an end-to-end AI creative studio

At Google I/O 2026, Google unveiled substantial updates to Google Flow and Google Flow Music, shifting its AI-powered tools from simple prompt-and-output platforms into comprehensive creative ecosystems. Elias Roman, senior director of product management at Google Labs, outlined how the updates address three critical creative demands: cross-modality to break down media-specific silos, defragmentation of fractured creative workflows, and enhanced precision with character consistency1

1

4

4

.

The transformation reflects Google's mission to build products that creative professionals return to throughout the entire creative process, rather than one-off generation tools. Roman emphasized that creators frequently navigate an expensive, disjointed ecosystem of single-purpose apps that disrupts their creative flow state1

1

. The updates aim to consolidate video generation, AI video editing, image creation, and music production into unified platforms.

Conversational agent brings agentic brainstorming to Flow

Google Flow now functions as a conversational agent powered by Gemini models, acting as an end-to-end creative co-pilot with full memory of past and current projects. This represents a fundamental shift from coin-operated generation to collaborative partnership4

4

. The Google Flow Agent serves as a sounding board for agentic brainstorming, helping creators decide on scene dialogue or plot direction, generate multiple scene versions for comparison, and execute batch edits1

1

2

2

.

This conversational capability enables creators to iterate naturally on their vision, with the AI maintaining context across sessions. The agent can help plan and reason through project development, making it particularly valuable for filmmakers working on complex narratives. The Google Flow Agent is available globally as of the announcement2

2

.

Gemini Omni Flash delivers precise video-to-video editing

Google integrated Gemini Omni Flash, a new generative model designed specifically for video creation, into the AI creative studio. Described as analogous to Nano Banana but for video production, Gemini Omni Flash unlocks precise, video-to-video conversational editing and robust character consistency, including users' own avatars1

1

. The model allows creators to blend real-world inspiration with generated content while iterating conversationally on outputs2

2

.

Crucially, Gemini Omni Flash preserves character identity and voice across scenes and generations, addressing a persistent challenge in AI video editing. Roman demonstrated this capability during a live demo featuring a decade-accurate 1980s-themed Times Square story with persistent constraints, including an Easter egg miniature pinscher dog appearing in every image. The model enables side-by-side video editing using natural language commands1

1

. Gemini Omni Flash is now available to all global Google AI subscribers1

1

2

2

.

Flow Tools enable vibe coding without technical expertise

Google launched Flow Tools, allowing creators to use natural language to instantly code and share custom tools or workflows. This feature, sometimes referred to as vibe coding, enables users to create specific utilities like video resizers or shaders without coding knowledge1

1

. AI agents generate the required code and build the applications based on natural language descriptions4

4

.

Creators can share their custom tools with other Flow users, building a collaborative ecosystem of bespoke workflows. Roman demonstrated custom tools including ASCII character video rendering and personalized avatar systems during his presentation1

1

. Flow Tools launched globally with some early tools already available for immediate use2

2

.

Native mobile apps bring AI creativity on the go

Google unveiled native mobile apps for both Google Flow and Flow Music, designed for flexibility and on-the-go creation. The apps coexist with web versions, enabling brainstorming and quick edits from anywhere1

1

2

2

. Flow's video editor launched first on Android in beta, with iOS coming soon, while Google Flow Music debuted on iOS first, with Android following later1

1

4

4

.

Google Flow Music gains granular editing and music video creation

Google Flow Music, which brings the Lyria 3 Pro music model to artists, producers, and songwriters, received its own substantial updates3

3

. The platform now offers granular editing capabilities, allowing musicians to edit single parts of songs—such as lyrics, translations, or beats—without affecting the rest of the track1

1

2

2

.

Creators can now generate covers of favorite tracks, keeping the melody and structure while changing the style—transforming pop playlists into lo-fi study versions, for example1

1

. Gemini Omni Flash integration enables AI music generation paired with music video creation, allowing users to conversationally guide styles, subjects, and scenes to match their track's narrative and pacing2

2

4

4

. These videos can be shared directly through the platform4

4

.

Building a dedicated product line for creative professionals

Google's updates signal a strategic shift toward building a dedicated product line entirely focused on AI creativity. Google Labs VP Elias Roman noted that while media generation tools initially functioned as coin-operated machines—users input prompts and receive outputs—the company now aims to create products that creative professionals use repeatedly throughout entire projects5

5

.

Google Flow was originally announced at Google I/O 2025 and has since expanded into over 140 countries3

3

. The platform's evolution from a simple video generation tool to a comprehensive AI creative studio reflects Google's commitment to supporting filmmakers, musicians, and creative professionals with integrated, AI-powered tools that streamline workflows while maintaining artistic control and vision.