Google Vids adds free AI videos: how Gemini Omni works

Google Vids and Gemini Omni: how the AI video feature worked in the September 2026 announcement, with account, language, and availability limits.

Red bicycle in a video scene, with successive positions suggesting movement created from a still frame.

Updated October 2, 2026. On September 23, Google announced that anyone with a Google or Google Workspace account could start creating videos with Gemini Omni 1.1 Flash in Google Vids without paying. The update expands a feature previously linked to paid plans. Usage limits apply, and the experience may vary by account. This guide explains what changed, how to try the tool on a computer, and which details to check before planning a project.

What Google announced

According to the official September 23, 2026 announcement, the “Create AI videos” workflow in Google Vids began offering creation with Gemini Omni 1.1 Flash without an initial charge for personal and Workspace accounts. Google also announced 1080p scene generation, upscaling of AI-generated clips, duration selection, scene extensions with continuity, and transitions. These features belong to the announced update; actual availability may depend on the account and displayed workflow.

The official generation help, checked again on October 2, now documents clips in 720p and 1080p at 24 frames per second, with a selectable duration of 3–10 seconds. Google’s announcement highlights 1080p in the Omni 1.1 workflow. Generation help still mentions eligible Workspace subscriptions, while the availability page lists personal accounts and allowances by plan. Documented resolution therefore does not guarantee that 1080p appears in every account, workflow, or export: check the options and resulting file in your interface.

What Google Vids and Gemini Omni are

Google Vids is Google Workspace’s video editor. It brings together scenes, media, text, audio, templates, and AI features in a project you can assemble and share. You can edit the sequence after generating a scene; the tool is more than entering a command and downloading a finished file. This helps anyone who needs a short explanation, training material, or presentation video and wants to adjust it before use.

Gemini Omni 1.1 Flash is the name Google uses in the announcement for the model applied to video creation in this Vids workflow. Do not treat “Omni,” “Veo,” and “Google Vids” as synonyms: Vids is the editing product; Omni is the model named in the update; Veo appears in earlier video-generation announcements and documentation. The July publication linking Omni to paid plans describes the situation then. The September announcement changes initial access to Omni 1.1 Flash in Vids.

Do not confuse this access with the video generator in the Gemini app either. The Gemini app help presents separate conditions and requires a paid Google AI plan for video generation in a personal account. The free route announced here is Google Vids, with its own limits.

Free for whom? Accounts, allowances, and availability

The announcement says signing in with a Google or Google Workspace account is enough to start without a charge. “Free” therefore means initial access to an allowance, not unlimited use. Vids’ availability page presents a combined monthly limit for AI clips and avatars for personal accounts, plus monthly credits; limits and features can change. Google AI plans and certain Workspace editions offer additional capacity. Check the allowance shown in your account before promising a number of videos to a client or team.

For work or school accounts, an administrator may control access. Personal accounts may have restrictions on some editing features. Documentation also states that several AI functions are available only in English. The announcement does not provide a specific guarantee for every region or confirm uniform availability in Brazil. If “Create AI videos” is missing, look for the creation panel and account availability information; do not infer an error solely from the nationality of the email address.

The documented workflow is for computers. Open Vids in a desktop browser for the test. We do not assume that complete creation works identically on a phone. For a professional project, also confirm your organization’s internal data and publication rules before inserting confidential material.

How to create your first AI video in Google Vids

We will follow one example: a small bicycle shop wants a short video advertising a preventive checkup. The goal is to show a red city bicycle moving, without asserting prices or terms the shop has not approved. This scope helps create a verifiable visual prompt and detect details invented by the model.

  1. Visit vids.new on a computer and sign in with your Google account. Open a new Google Vids project.
  2. Look for Create AI videos. The interface may display a creation panel with prompt fields and variable controls. If the option is absent, check account availability and administrator controls in Workspace.
  3. Describe the scene’s subject, setting, action, framing, and lighting. Begin with one scene, generate it, and assess the result before using the allowance on new versions.
  4. During creation, select a duration of 3–10 seconds according to Google’s help. If the interface offers extension, use it only when you need to continue the same action; check that the bicycle, setting, and direction of movement remain consistent.
  5. Arrange the scenes in the editor. Use a template if it helps the format, adjust text and audio, and include a closing call to action you have reviewed. Transitions can improve the passage between scenes but do not fix a confusing script.
  6. Review the entire video, including the beginning, continuity, text, audio, and final frame. Check the exported file’s resolution and account usage conditions before sharing.

A first prompt could be: “Short, realistic video of a red city bicycle passing along a quiet street at dawn, a side camera tracking the movement, soft light, no text in the image, no visible brands; keep the bicycle’s shape and color consistent.” If wheels are deformed or the color changes, request another shot with less movement or simpler framing. Specificity helps but does not guarantee correctness.

For a second scene, try: “The same red bicycle arrives at a bright workshop; close shot of a mechanic’s hands checking the front brake, natural movement, no logos, no generated text.” Before joining the scenes, check whether they actually look like parts of the same video. The announced extension aims for continuity, but human review remains indispensable.

What changed in practice

The first change is the entry point. People without a paid plan can now try the Omni 1.1 Flash workflow inside Vids, subject to their account allowance. That makes testing a video idea before purchasing more capacity feasible. The second is the announced set of controls: duration selection, the ability to extend scenes, and transitions in an editor. The third is the product context: a generated clip enters a project that can combine templates, edited scenes, and other elements.

Google says generated videos include an imperceptible SynthID digital watermark. It is not necessarily a visible badge in a corner and does not remove the need for editorial transparency when artificial content could confuse viewers. When publishing a commercial example, identify the simulation where necessary and check rights covering images, music, brands, and people used in the project.

Google’s help documents 720p and 1080p at 24 fps for generated Vids clips. It also instructs users to choose 3–10 seconds per new or edited clip. The announcement emphasizes new 1080p scenes and scene extension. This describes feature options, not a promise of 1080p in every account or unlimited final duration: confirm what your interface allows and check the exported file.

Where to use it and where to be cautious

Vids makes sense for prototypes, internal communication, short demonstrations, presenting an idea, and visual content that will be reviewed in the editor. A team can produce a draft, discuss the sequence, and redo only the problematic scene. A teacher can illustrate a simple concept if they check the result’s accuracy. A small business can try styles before recording a final version.

It may not be the best choice when every detail must be faithful, such as a medical procedure, a product with exact specifications, or a real person whose appearance must not change. Video generation may invent elements, replace objects between frames, or depict movement poorly. Obtain permission to use faces, brands, or third-party material. Do not use the result as documentary proof of an event.

The free allowance also calls for planning. Write the script and prompts outside the generator, decide what must appear, and reserve attempts for real corrections. A longer video may require several clips and manual editing. Free access works well for validating a concept; recurring volume may require another plan or production workflow. Compare the concrete need with the limit shown in the account without treating an allowance published today as a permanent promise.

Google Vids and other AI video tools

For a broader view of AI-assisted audiovisual production, read our guide to creating videos with AI. It covers the general process of scripting, generation, and editing. Here the focus is the specific September 2026 news: opening the Gemini Omni 1.1 Flash workflow in Google Vids and the conditions to check before using it.

To compare options without mixing products, begin with the task. If you already work with Workspace documents and presentations, Vids may simplify moving to a video editor in the same environment. If you only need separate clips, check the allowance and export quality. If the work requires strict visual identity, long scenes, or frame-by-frame control, test a representative script before choosing the tool. A model’s promise does not replace testing your own workflow.

The AI news hub brings together recent changes worth following. For Vids, account and regional availability deserve revisiting. Help already documents 720p and 1080p, but generation help and the announcement still use different eligibility wording. Consult the option displayed in your account before planning production; documentation, allowances, and availability can change.

Frequently asked questions

Is Google Vids with Gemini Omni free?

There is initial access without a charge for Google or Google Workspace accounts according to the September 23 announcement. This access has an allowance and may vary by account type. Paid plans offer additional capacity. Check the balance and available features in the interface itself.

Do I need a Gemini subscription to create video in Vids?

The Vids announcement says you do not need to pay to start using Omni 1.1 Flash in this product. That does not change video-generation conditions in the Gemini app, which has its own documentation and plans. Open Google Vids to try the workflow described here.

Is the video always exported in 1080p?

Not necessarily. Help lists 720p and 1080p at 24 fps, and the announcement highlights 1080p for new Omni 1.1 scenes. The actual option may vary by account and workflow; check the exported file’s resolution before promising 1080p.

Can I create it on a phone?

Google’s described route uses Vids on a computer. To follow this tutorial, use a desktop browser. We do not assume feature parity with mobile.

What is a clip’s maximum duration?

Google’s help instructs users to select between 3 and 10 seconds for a new or edited clip. Scene extension is also available, subject to availability and account allowance. A clip’s range is not the maximum duration of a complete project; check the options in your interface.

Sources and verification date

Information checked again on October 2, 2026 in Google’s announcement, Google Vids availability, video-generation help, and Gemini app help. Help began documenting 720p and 1080p, 24 fps, and selectable durations of 3–10 seconds. Eligibility and allowances still require checking account type and interface.