Google has added new AI agents, editing tools and mobile apps to Google Flow and Google Flow Music, extending its creative software for video, image and music work.
The update brings Gemini Omni Flash to Google Flow, along with a new Google Flow Agent for project planning and editing, and a tools feature that lets users build custom workflows with natural language. In Google Flow Music, Google has introduced section-by-section song editing, full-track covers and music video creation with Gemini Omni.
Flow, originally introduced for filmmakers, is now available in more than 140 countries. Google has since expanded it into a broader creative studio for video and image generation and editing, while Flow Music brings its Lyria 3 Pro music model to artists, producers and songwriters.
Video tools
At the centre of the Flow update is Gemini Omni Flash, a model Google describes as able to create from different forms of input, starting with video. It combines Gemini with Google's generative media models and is designed to support conversational iteration as users mix real-world material with generated content.
Omni Flash also improves character consistency so identity and voice remain stable across scenes. The model is available in Flow to Google AI subscribers worldwide.
Google is also rolling out Google Flow Agent to all Flow users globally. The agent is intended to assist with brainstorming, creation and editing, including dialogue development, plot suggestions, producing multiple variations at once and batch-editing assets so changes carry across a project.
Once material is created, the agent can organise assets into collections and rename them. The feature is built on Gemini models and designed to work under user control.
Custom workflows
Another addition is Google Flow Tools, which lets users create bespoke tools and workflows through natural language prompts rather than code. Creators can build utilities such as image editors, video resizers and custom shaders, then share them with other Flow users for reuse or modification.
Existing tools are available to all Flow users, while Google AI subscribers can create and remix them. Google cited early access partner László Gaal, who made a tool called "pixelBento" for post-processing effects including lo-fi and glitch styles.
It also highlighted tools from Kat Zhang and metapuppet, presenting the feature as a way for users to adapt the software to their own production methods rather than rely on a fixed set of editing options.
Music changes
For Flow Music, the update focuses on more precise editing and broader video output. Users can now highlight individual parts of a song and alter only those sections, such as rewriting or translating lyrics, changing a beat drop, or sampling one section and extending it in a new direction without affecting the rest of the track.
Google has also added a covers feature that lets users transform the style of complete songs while keeping their original melody and structure. It gave the example of turning a playlist into a lo-fi study version.
Gemini Omni is also coming to Flow Music for music video creation. Users can work conversationally with the agent to direct videos and guide style, subject matter and scenes to align with a track's narrative and pacing. This feature is available to Google AI subscribers.
Mobile access
Alongside the web products, Google is launching mobile apps for both Flow and Flow Music. The web versions remain the main platforms for full access to features, while the apps are intended to support creation away from a desktop setup.
For users aged 18 and over, the Flow app is available in beta on Android, while the Flow Music app is available on iOS. Android support for Flow Music and iOS support for Flow are due later.
The broader update shows Google continuing to fold more of its generative AI models into products aimed at creative work, as major technology groups compete to offer software for video production, music creation and editing. By linking Gemini models with video and audio tools, Google is trying to make those systems useful not only for generating material from prompts, but also for managing revisions and project organisation.
Millions of creators across a range of disciplines have used Google Flow and Google Flow Music over the past year, according to the company.