How to Use Google Flow’s Six New Creator Tools

Google published six new Google Flow Tools on 23 September 2026. The gallery is not a generic prompt box. Each tool is a reusable workflow built with working creatives in sound design, architecture, and digital content.

You can run the published tools on a standard Flow account. Google AI subscribers can also remix a tool or describe a new one in plain language. This guide walks through what each tool does, how to open the gallery, and how to duplicate a workflow without writing code.

What Google Flow Tools are

Google Flow is Google Labs’ workspace for generating and editing video, images, and related media. Flow Tools sit beside that workspace. You describe a job once. Flow turns the description into a shareable interface with inputs, sliders, and export steps.

Google first opened custom tools to all Flow users in May 2026, when Gemini Omni arrived in Flow. Anyone can run tools already in the gallery. Creating and remixing tools is limited to Google AI subscribers. Features can still vary by plan, region, and whether you are on web or mobile.

The September drop is narrower. Google Labs worked with people in architecture, sound design, and digital content, then shipped six tools those teams actually use.

Open the Tools gallery

1. Sign in at flow.google with the same Google account you use for Flow credits.
2. Open or create a project so the left navigation is available.
3. Select Tools. That opens the Tools gallery Google and partner creatives publish into.
4. Search by name (for example CaptionCast) or open a shared tool URL from the official announcement.
5. Pull a clip or image from your project into the tool, or generate a new asset inside it.

If you only need a one-off result, stop there and export. If you want a variant, use Remix at the top of a tool you already like. Remix copies the logic so you can change prompts, labels, and sliders. To start from nothing, go back to the gallery homepage and choose Create Tool, then describe the editor you want.

Official shared links for the six tools live on the Google Labs announcement. Bookmark those URLs if your team will reuse the same workflow.

The six tools, and when to use each one

1. Mondo Sónico

Creative director Ricardo Villavicencio and sound designer Sebastián Carvallo built Mondo Sónico after years of needing more flexible beds under picture.

The tool generates background ambience, foley, and contextual effects on separate, editable tracks that stay in sync. Export individual stems into a digital audio workstation instead of flattening everything into one mix. Use it when you have picture locked and need a first-pass sound bed you can still split later.

If you already generate music in Gemini, pair this stem workflow with the track-building steps in our Lyria 3.5 music guide. Flow Music and Mondo Sónico solve different jobs: songs versus scene audio.

2. CaptionCast

Digital storyteller Jay Pirabakaran built CaptionCast to remove hand-timed subtitles. One pass transcribes speech, styles the type, and animates multilingual captions.

Use it on short-form cuts where every extra export round costs you a posting window. Check the transcript before you publish. Automatic captions still miss names, product terms, and language switches.

3. ThumbnailForge

Also from Pirabakaran. ThumbnailForge takes one image, a headline, and a prompt, then renders cover art sized for social platforms.

Feed it a still from the same project as the video so color and subject match. Write the headline first. Vague prompts produce generic posters. Specific platform and mood notes produce usable covers.

4. Surface

Interior designer and technologist Vojtek Morsztyn built Surface for the slow part of architectural ideation: sourcing and testing materials.

The tool generates custom textures, from minimal patterns to Art Deco tiling, and maps them onto 3D walls, ceilings, and floors in real time. Use it for mood boards and client options, not as a substitute for a specification sheet or a physical sample.

5. CollageMotion Pro

Filmmaker and science communicator Hashem Al-Ghaili built CollageMotion Pro for mixed-media documentary looks that usually take hours of manual compositing.

Give it a text prompt. It returns animated collages you can scale across a series. Keep source rights in mind. If the collage includes real photographs or archival stills, confirm you can use those assets before you ship the cut.

6. SwissFlow Studio

Al-Ghaili also shipped SwissFlow Studio. It turns a written script into Swiss-style motion graphics and reduces manual keyframing.

This is the tool for title cards, explainer beats, and marketing sequences that need clean type more than photoreal video. Write the script in short lines. Long paragraphs produce crowded frames.

Remix a tool without starting over

Google’s own Flow team shows the loop in the Find Your Flow series: open a gallery tool, drop in a clip, then remix when one preset is close but not right.

1. Open the tool and run it once so you can see the default controls.
2. Select Remix. Flow copies the tool into your account.
3. Add a short instruction in the agent chat, such as a new caption language, a fisheye preset, or a different export layout.
4. Generate again and compare against the original output.
5. Save a version that works so it appears in your project gallery.

Keep remixed tools narrow. A “do everything” editor is harder to debug than a caption tool that only handles three languages you actually publish in.

Build a tool from a sentence

On the gallery homepage, choose Create Tool. Describe the editor as if you were briefing a junior teammate.

Good briefs name the input, the transform, and the export:

• “Take a 16:9 product clip, add lower-third captions in English and Spanish, and export a 9:16 crop.”
• “Generate three room textures from a photo of a wall and map them onto a simple box room.”
• “Build a shader rack with grain, bloom, and a teal-and-orange grade I can apply to Flow clips.”

Flow shows an overview of the generated controls in the agent chat. Test immediately. If a slider does nothing, say so in the same chat and ask for a fix. Google AI subscribers can share a finished tool with other Flow users, who can then remix it again.

Practical limits

Credits still apply. Gallery tools consume the same Flow generation budget as a normal project render. Plan stems and collage batches instead of firing ten variations at once.

Availability is not identical everywhere. Google has expanded Flow and higher AI plans across many countries, but product notes still flag differences by subscription tier, platform, and region. Age restrictions listed on Flow pages also apply.

Do not treat generated audio, captions, or architectural textures as final legal or construction documents. Review stems, transcripts, and material maps the same way you review any first draft.

If your day job is writing in Workspace rather than cutting picture, Gemini in Gmail is a better starting point than Flow. See how to use Gemini in Gmail for that path.

A short production checklist

• Open Flow, create a project, then enter Tools before you generate anything expensive.
• Start with a published tool that matches the job instead of inventing a new editor on deadline.
• Export stems from Mondo Sónico and captions from CaptionCast as separate files.
• Remix only after one clean baseline output exists.
• Keep shared tool links in the shot list so editors land on the same version.

Conclusion

The six September tools are opinionated. They encode how specific creatives already work: split audio, animated captions, social covers, mapped materials, collage motion, and Swiss type.

Use the gallery when the job matches. Remix when you need one extra control. Build from scratch only when no listed tool is close. That order saves credits and keeps the interface small enough to trust on a real deadline.

Sources

• See six new tools in Google Flow — Google Labs, 23 September 2026
• Introducing Gemini Omni for Google Flow and Flow Music — Google Labs
• How to use Tools in Google Flow — Find Your Flow, YouTube
• Google Flow