Google Shopping’s AI Try On Goes Live Across the Philippines

Google’s AI Try On feature is now available in the Philippines, letting shoppers upload a photo of themselves and virtually preview clothing and footwear on their own body before checkout. The rollout, announced June 2, covers seven Asia-Pacific markets simultaneously and draws on the company’s Shopping Graph, a catalog of more than 50 billion product listings from retailers worldwide.

The feature first ran as a limited US pilot in July 2023. Its arrival in the Philippines comes as the country’s e-commerce market, where fashion and apparel generate the largest category share, has reached a scale where visual AI tools carry real commercial weight.

Seven Markets in One Go

Google pushed Try On to Hong Kong, Indonesia, Malaysia, the Philippines, Singapore, South Korea, and New Zealand in a single announcement, the largest single geographic expansion by country count since the tool launched. Earlier additions came in smaller batches: Australia, Canada, and Japan in October 2024, with the UK and India following afterward.

The timing follows Google I/O 2026, held in May, where the company announced a broader AI shopping push. New features included Universal Cart, a Gemini-powered cart that spans Search, YouTube, and Gmail and actively tracks price drops on saved items, and an upgrade of AI Mode in Search to run on Gemini 3.5 Flash. The Try On feature has its own separate rollout track but fits the same direction: compressing the steps between a search result and a purchase decision.

Our goal is to be the best dressing room in the world and the place. Instead of having to walk into seven different shops, you can take all of the things that you love and see them on yourself in one place.

Dominique M, Google’s senior product manager for GenAI for shopping, made those remarks during a media briefing on the APAC launch.

To access the feature, shoppers need a Google account with Search Services History and Personalized Recommendations turned on and must be at least 18 years old. The rollout across the Philippines will take place over the coming weeks as Google activates it for more clothing categories and retailers.

How a Single Photo Generates a Fitting Room

The Model Under the Hood

The feature runs on a custom image generation model built specifically for fashion, developed in collaboration with Google DeepMind, the company’s AI research division. The model was trained on pairs of photos showing people in different poses wearing the same garment, teaching it how different fabrics fold, stretch, cling, and drape across varying body types and postures.

Before the personal-photo upload option existed, the feature worked through a library of approximately 80 real models spanning a range of body types, sizes from XXS to 4XL, and ethnic backgrounds. That model gallery remains available alongside the photo upload pathway. Google cited at the original US launch that 42% of shoppers don’t feel represented by standard model images, which shaped the diversity of that initial library.

No 3D avatar is created. The AI generates an entirely new image showing the chosen garment on the user’s body, not a static clothing layer placed over a photo. “Accuracy matters when you’re shopping,” Dominique M said in the same media briefing, noting that the system draws on DeepMind’s 3D modeling capabilities and training data from millions of products in the Shopping Graph.

Using the Feature Step by Step

1. Search for an apparel item on Google Search, Google Shopping, or Google Images
2. Look for the “Try it on” icon on eligible product listings
3. Upload a full-length photo: head to toe, good lighting, clean background, hands visible at your sides
4. The model generates a preview within seconds
5. Save or share the result; tapping the shop button goes directly to the merchant’s site for checkout

Supported categories at launch include tops, bottoms, dresses, shoes, and bags and accessories. Color variants appear automatically when the listing includes them. Google Shopping’s Try On help documentation notes the feature is not available for Shopping Ads; products listed under “Sponsored” do not show the “Try it on” button.

US users have an additional option: Nano Banana, Google’s Gemini 2.5 Flash image model, can build a full-body digital version of the shopper from a single selfie, which then serves as the default try-on photo. That pathway is US-only and is not part of the Philippine launch.

Fashion Already Leads Philippine Online Sales

• USD 28.0 billion: Philippines e-commerce market size in 2025, per IMARC Group, a market research firm
• 27%: Fashion and apparel’s share of gross merchandise value (GMV) on Shopee Philippines in FY 2025, per Cube’s Philippines e-commerce market data
• 57%: Share of Philippine e-commerce volume transacted on mobile devices in 2024, per industry research
• 25%: Average global return rate for online apparel purchases, per Eightx’s 2026 e-commerce return rate analysis

Fashion and apparel lead Philippine online retail across platforms. TikTok Shop generated $2 billion in GMV in the Philippines in Q4 2025, overtaking Lazada as the second-largest platform by quarterly sales, according to Cube. Fashion and apparel account for 26% of TikTok Shop’s Philippine GMV, and live selling, where hosts model clothing for mobile viewers in real time, has become a primary discovery format for fashion purchases in the country.

Mobile-first browsing compounds a persistent visualization problem. Choosing clothing on a phone from a standard product image leaves fit, silhouette, and drape entirely to guesswork. Returns measure that gap: apparel return rates run at more than twice the level seen for electronics or beauty products, with size and fit issues driving the majority of those decisions, according to Eightx’s 2026 analysis.

Google’s own support documentation notes the feature “aims to provide a helpful visualization, but it’s an approximation and may not reflect body shape, personal features, or product details with perfect accuracy.”

What Merchants Need to Know

Participation is automatic. Any merchant with a shopping feed who is eligible to show free listings on Google is opted into Try On without any additional steps. The AI uses product images already in the Merchant Center feed to generate try-on renders, so image quality directly affects output. Google’s Merchant Center Try On guidelines require each product image to show a front-facing model in a simple pose wearing a single garment.

A Google representative confirmed at the APAC briefing that the feature has no impact on product rankings or merchant visibility. “We don’t change anything about ranking or decisions,” the representative said. “This is only available on our organic listings. We are just a button on every product that you can use to try things on.”

Merchants who opt out only remove the button from their own listings. If another merchant sells the same product and stays opted in, the Try On button continues to appear on that merchant’s listing.

Privacy Guardrails and Hard Limits

Uploaded photos are processed to generate a try-on render and are not retained. Google does not collect or store any biometric data during the experience, and the uploaded photo is not used for AI model training or shared with other Google products, services, or third-party affiliates.

Specific photo requirements apply. Full-length photos need good lighting, a clean background, and the user standing with hands visible. Overly baggy clothing, photos taken too close to the face, shots with the subject sitting or crouching, and images taken from too far away all degrade the render quality. Uploading photos of children is not permitted, and all images must comply with Google’s Generative AI Prohibited Use policy.

Generated images can contain errors in body shape, clothing details, or personal features. Users who spot inaccuracies can flag them through the feedback link at the bottom of the try-on screen.

A Bigger Play for the Purchase Moment

The Philippine launch fits a larger Google push to make Search itself a shopping destination. At I/O 2026, the company announced Universal Cart, described in Google’s I/O 2026 product announcements as a hub where users can add items while browsing Search, chatting with Gemini, watching YouTube, or reading Gmail, with the cart scanning for price drops and stock alerts in the background. Agentic checkout in the US already lets Gemini complete purchases via Google Pay once a shopper confirms price and item details.

The competition has moved quickly around the same ground. Earlier this year, OpenAI unveiled “Instant checkout,” letting users complete purchases directly inside ChatGPT through a commerce protocol co-developed with Stripe, with merchants retaining control over payments and fulfillment. Perplexity launched its own virtual try-on tool in November 2024 for Pro and Max subscribers. Valuates, a market research firm, projects the global virtual try-on market will grow from $5.8 billion in 2024 to $27.7 billion by 2031, driven by smartphone adoption, AI improvement, and persistently high return rates in fashion categories.

Try On sits at the top of that funnel, at the moment a shopper is considering a purchase but hasn’t committed. In the Philippines, where fashion and apparel drive the largest share of online retail volume and most of that browsing happens on mobile, the rollout lands in a market built for exactly this kind of visual-first commerce tool.

The feature continues rolling out across the Philippines over the coming weeks.