Google is expanding its AI tools for preparing advertising campaigns: automating search ad localization, enabling the creation of product images in Asset Studio, and introducing Gemini Omni 1.2 Flash for generating videos up to 15 seconds long. For advertisers, this means an opportunity to adapt campaigns to new markets faster and test more creatives. However, feature availability, supported languages, and usage rules vary.
Key Takeaways
- AI Campaign Localization creates a separate localized copy of a campaign without changing the original.
- The beta tool supports translation from US English into 10 language variants. Ukrainian is not among them.
- Google Ads lets advertisers translate ad copy and keywords and configure landing page localization.
- Asset Studio creates advertising images from photographs of physical products, including those from Merchant Center.
- On November 5, 2026, Google introduced Gemini Omni 1.2 Flash with support for videos lasting 3–15 seconds.
- Generating 720p video through the Gemini API costs $0.07 per second — 30% less than the previous model.
AI Campaign Localization: launch ads in a new language without manual copying
Google Ads is testing AI Campaign Localization, a tool for adapting advertising campaigns to other languages and markets. Its main advantage is that advertisers do not need to transfer each ad, keyword, and extension manually.
The system creates a new, independent campaign based on an existing one. The original campaign remains unchanged, so the localization can be reviewed separately from the current advertising campaign.
The tool works with several components:
- Ad copy: headlines, descriptions, sitelinks, and callouts.
- Keywords: translation into the selected campaign language.
- Images containing text: AI removes text from the image and adds the translated message as a separate ad text element.
- Landing pages: dynamic translation via JavaScript or the use of existing localized URLs.
- Campaign settings: copying budgets and bids from the source campaign.
One important detail: Google does not convert budgets using exchange rates or adjust them to advertising costs in the new market. If the original campaign has a $100 budget, the localized copy inherits the same amount in dollars until the advertiser changes it.
Which languages does AI Campaign Localization support?
According to Google’s official documentation, the source language for localization currently must be US English.
Ten target language variants are available:
- Dutch;
- French and Canadian French;
- German;
- Italian;
- Polish;
- Brazilian and European Portuguese;
- Spanish for Spain and Latin America.
These are 10 language variants, not 10 different languages. Ukrainian is not currently supported. Google plans to expand the list by the end of 2026 but has not named any specific additional languages.
The documentation describes the feature primarily for Search campaigns. Google also mentions expanding support to Performance Max from September 2026, although actual availability needs to be checked in each advertising account.
How to review translations before launching a campaign
Google has included a dedicated review stage. After generation, the advertiser receives a table showing the original text, its translation, a back-translation into English, and notes from the system.
Back-translation helps identify cases where AI has changed the meaning of an offer, misrepresented a product benefit, or distorted an advertising message.
Users can manually edit the text and add instructions, such as keeping a brand name untranslated or following the company’s official communication style.
By default, the localized campaign is offered for publication with a Paused status. This allows the review to be completed before any spending begins.
For international campaigns, this matters more than translation speed alone. Automated localization can accelerate preparation but does not guarantee that an advertising offer will fit the cultural context, legislation, and search habits of the new market.
Landing page translation requires separate configuration
AI Campaign Localization does not automatically create a complete, separate localized website.
Google offers two main approaches. The first is to install a JavaScript snippet on the website that dynamically translates landing page text for users arriving from ads.
The second is to use your own localized pages and replace the URLs in the campaign manually or in bulk using a find-and-replace rule.
For ecommerce, financial products, and regulated categories, the second approach may be more practical when local delivery and return terms, legal disclosures, or other information are needed that cannot be accurately adapted through translation alone.
Asset Studio creates advertising images from product photos
The second area of updates is product image generation directly in Google Ads.
In Asset Studio, an advertiser can upload one or more photos of a physical product and describe the desired setting. For example, a standard studio photo of a backpack can become a lifestyle image for a travel campaign.
Google recommends providing photos from different angles so the model can reproduce the product’s shape and features more accurately.
Images can be selected from the advertising account’s asset library, uploaded manually, or chosen from a connected Merchant Center.
After generation, the result can be edited, the text prompt changed, or the finished image added to an ad or asset group.
UAGEEK previously examined how Asset Studio automates asset preparation for Performance Max. The new product photo capability builds on that process: advertisers can create several visual concepts from an existing catalog.
Who can access image generation in Google Ads?
The feature is rolling out gradually. Google warns that even advertisers who meet all the requirements may not yet see the relevant tool.
Google sets several conditions for manually generating images from text prompts:
- the advertising account has been active and running ads for at least 30 days;
- the account has a history of compliance with advertising policies;
- the advertiser has not completed political advertising verification;
- the business does not belong to certain sensitive categories;
- the interface and text prompts use one of the supported languages.
The supported prompt languages listed are English, Dutch, French, German, Italian, Japanese, Portuguese, and Spanish. Ukrainian is not on this list.
Some automatically suggested images may also be available to newer accounts, but this does not mean those accounts have full access to generation using their own prompts.
Gemini Omni 1.2 Flash: videos up to 15 seconds and cheaper generation
On November 5, 2026, Google introduced Gemini Omni 1.2 Flash, an updated video generation model. It supports creating clips from 3 to 15 seconds long, in one-second increments.
The previous Omni 1.1 version supported clips of up to 10 seconds. The new limit provides more room for short advertising stories, product demonstrations, and vertical video creatives.
Google has also improved support for references. The model can take into account images of characters, products, and styles, as well as short video clips and audio cues.
According to the company, this helps preserve character appearances, clothing details, product packaging, and brand colors more accurately across frames.
Another change is multi-step editing. After creating a clip, users can ask the model to change the background, lighting, weather, product, or aspect ratio without starting the entire process over.
For advertising teams, this could reduce the time needed to create variations of a single video for different audiences and placements.
How much does Gemini Omni 1.2 Flash cost?
Google has reduced the cost of generating 720p video by 30% compared with Gemini Omni 1.1 Flash.
The announced price through the Gemini API is $0.07 per second of 720p video.
At this rate, a 15-second clip would cost approximately $1.05 for one generation. This calculation is based on the published price and does not include potential retries, edits, or other costs.
The model is available through the Gemini API and Google Cloud. Google has also announced its use in Flow, Google Vids, the Gemini app, and the company’s advertising tools.
However, the API price does not mean that every generation within Google Ads or other products is billed in the same way. Terms depend on the specific service.
UAGEEK has already covered the integration of Gemini Omni into Google Ads. Version 1.2 Flash expands the model’s capabilities, particularly video duration, reference handling, and editing.
AI creatives still need to be reviewed before launch
Faster generation does not eliminate risks for advertisers. Google explicitly warns that AI-generated images may contain inaccuracies and that generated assets do not receive automatic approval from ad moderation.
It is particularly important to check the product’s appearance, packaging text, logos, product specifications, and advertising claims.
Localized campaigns also introduce language-related risks: mistranslated promotional terms, poorly worded keywords, or a mismatch between the ad and the landing page.
AI content labeling requirements must also be considered. Google notes that in some jurisdictions, ads containing artificially created or edited materials may require appropriate disclosure.
These rules are discussed in more detail in UAGEEK’s article on AI creative labeling on Google, Meta, and TikTok.
What performance marketing teams should do
- Check tool availability. Look for AI Campaign Localization and Create product images in your Google Ads account.
- Choose a test market. Start with one campaign and one supported language instead of localizing the entire account at once.
- Review translations. Check keywords, offer terms, brand names, and landing pages.
- Prepare high-quality references. Use clear product photos from several angles.
- Monitor video generation costs. Consider not only the cost of the final clip but also the number of attempts and edits.
- Test the results. Compare AI creatives with conventional ones using CTR, conversions, CPA, and other business metrics.
The main practical effect of these updates is less manual work during asset preparation. However, advertising effectiveness must still be demonstrated through campaign results, not the number of generated variations.
Conclusion
Google is gradually bringing localization, image creation, and video generation together within a broader ecosystem of AI advertising tools. AI Campaign Localization simplifies campaign preparation for other languages, Asset Studio helps scale product visuals, and Gemini Omni 1.2 Flash enables longer videos at a lower price.
However, these are three distinct tools with separate access conditions. For advertisers, the priority is to check supported languages, account restrictions, output quality, and the actual economics of using AI.
Sources for verification
Sources for verification
Google Ads Help — About the AI campaign localization tool for Search campaigns (beta). Official documentation covering campaign localization, supported languages, budget copying, landing page translation, and pre-launch review.
How AI Campaign Localization works
Google Ads Help — About generated images in Google Ads. Description of product image creation in Asset Studio, use of Merchant Center, advertising account requirements, and AI generation limitations.
Advertising image generation in Google Ads
Google Blog, November 5, 2026 — Gemini Omni 1.2 Flash. Official model announcement: videos up to 15 seconds, multi-step editing, reference support, a price of $0.07 per second at 720p, and a 30% cost reduction.
Gemini Omni 1.2 Flash announcement