AR Product Try-On for Shopify: Setup Guide (2026)

Why AR Try-On Finally Matters for Shopify Merchants in 2026

Products with 3D or AR media on Shopify convert at roughly 94% higher rates than standard product listings, according to Shopify's own ROI research on AR shopping. That's not a rounding error — it's the biggest single swing a product page can make without redesigning the theme.

For years, AR felt like a novelty reserved for IKEA, Warby Parker, and luxury watch brands with deep pockets. In 2026, the tooling has caught up. An iPhone with a LiDAR sensor can scan a product in 90 seconds. AI services generate a usable 3D model from four photos. Shopify ships native model-viewer support in every modern theme. The friction that kept AR out of reach for solo merchants is gone — and the merchants who move first are locking in conversion advantages that show up immediately in their dashboards.

This guide walks through the full setup for AR product try-on on Shopify: the file formats you actually need, how to generate 3D models without a studio, which apps fit which verticals, how to integrate model-viewer into a custom theme, how to handle performance, and — maybe most importantly — when not to bother with AR at all. If you sell apparel, eyewear, jewelry, furniture, or home goods, there's a version of this workflow that pays for itself inside a month. If you sell consumables or commodities, you'll learn why you should skip it. Either way, start with a grounding in modern theme-design fundamentals before touching the technical layer.

How Shopify Handles 3D and AR Natively

Before you install a single app, understand what Shopify gives you out of the box. Every plan — from Basic to Plus — supports 3D model uploads on product pages. The rendering pipeline uses Google's open-source <model-viewer> web component, which Shopify serves from its CDN alongside your theme assets.

The native feature set

• 3D model uploads in the product media gallery, alongside images and video
• Automatic device detection — iOS visitors see an "AR Quick Look" button, Android visitors get Scene Viewer, desktop gets an interactive 3D rotator
• No app required for the baseline 3D + AR view-in-your-space experience
• Shop app integration — 3D models also render in Shop's AR previews, the Shopify-owned consumer shopping app
• Theme-level rendering via the model_viewer_tag Liquid filter

What Shopify doesn't give you natively

The native tools handle view-in-your-space AR beautifully — perfect for furniture, décor, and large items a customer wants to place in a real environment. What they don't handle is try-on AR, where the camera tracks the user's face, hands, or body. For rings, glasses, necklaces, hats, shoes, and apparel, you'll need a dedicated try-on app (we'll cover picks below).

The practical rule: native Shopify AR answers "does this fit in my space?" Dedicated try-on apps answer "does this fit on me?" Most merchants end up using both.

The File Formats You Actually Need: GLB and USDZ

Shopify's 3D pipeline is built around two formats, and understanding the difference saves hours of rework.

GLB (GL Transmission Format Binary) is the web standard. It packages geometry, textures, materials, and animation into a single binary file, and it's what model-viewer renders on desktop and Android. If you upload only one file, make it GLB.

USDZ (Universal Scene Description Zipped) is Apple's native AR format. iOS 12+ devices open USDZ files in AR Quick Look — the full-screen AR experience with shadows, occlusion, and the "tap to place" UX that feels polished by default. Without a USDZ, iOS users get a degraded AR experience.

The dual-file rule

Upload both formats to every product. Shopify serves the right one based on the visitor's device automatically. You don't write any conditional logic — the <model-viewer> component handles it.

Compression matters

Apply Draco mesh compression to every GLB before upload — it typically cuts file size by 70% with zero visible quality loss. For USDZ, bake textures at 2K resolution max (4K is wasted bandwidth on a phone screen). A well-optimized apparel item should land around 2-3 MB GLB and 8-12 MB USDZ. Anything larger and you'll feel it in product-page load time.

How to Generate 3D Models Without a Studio

This used to be the blocker. Not anymore. You have three realistic paths depending on your budget, volume, and product category.

Path 1: iPhone LiDAR scanning (best for home goods, furniture, décor)

Any iPhone Pro model from the 12 onward has a LiDAR sensor. Download Polycam or Scaniverse, walk around the product twice, export as GLB. You'll get a usable model in under five minutes. Quality is excellent for opaque, textured objects and mediocre for glass, chrome, or highly reflective surfaces.

Strong fits: ceramics, wood furniture, pottery, planters, candles, upholstered items, sneakers, bags.

Weak fits: crystal glassware, mirrored jewelry, anything transparent.

Path 2: AI-generated models from photos (best for apparel and rigid goods)

Services like Kaedim and Polycam's photogrammetry mode generate 3D models from 4-20 photos. Kaedim specifically targets ecommerce — you upload product photos, their pipeline returns a production-ready GLB in hours. Cost is typically $10-50 per model. Shopify's own team covers the landscape in their AI 3D model generation tools roundup, which walks through quality trade-offs across providers.

Path 3: Dedicated 3D scanners or studio services (best for jewelry, watches, high-value SKUs)

For products where a slight mesh imperfection tanks perceived quality — fine jewelry, luxury watches, electronics — you'll want a professional capture. Services like CGTrader's on-demand 3D modeling and Shopify Experts specializing in 3D charge $50-300 per model with a 3-7 day turnaround. If you have 5,000 SKUs, this doesn't scale; if you have 50 hero products, it's the right move.

Volume math

At 200 SKUs and $30/model via AI generation, you're at $6,000. If AR lifts conversion even 30% on those pages (well below the Shopify-reported average), the payback period is measured in weeks for most established stores. That calculation breaks for stores under $50K/year in revenue — which is the first honest warning we'll dig into later.

Best Shopify AR Apps for Try-On Experiences

Native Shopify covers view-in-your-space. For true try-on — where AR tracks faces, hands, or bodies — you need a specialist app. Here are the ones worth shortlisting in 2026, organized by category fit.

Camweara — the generalist try-on platform

Camweara supports jewelry, eyewear, watches, hats, apparel, wigs, and nail polish in a single app. Its hand-tracking is among the best in the category, and it includes a Ring Sizer feature that measurably reduces size-based returns. If you sell across multiple accessory categories, this is the least-painful choice.

Poplar AR — 3D product viewer from Poplar Studio

Poplar AR/3D Product Viewer pairs a solid 3D viewer with optional try-on modules. The team also offers a concierge 3D modeling service — useful if you don't want to run the Polycam or Kaedim workflow yourself. Poplar's own team wrote a solid getting-started guide to Shopify AR worth reading before installing.

ARitize 3D by Nextech AR Solutions

ARitize 3D uses AI to convert existing 2D product images into 3D models, then serves them back through a viewer component. Lower friction than scanning, but quality varies by product — test it on your weirdest SKU first, not your cleanest one.

Augment — furniture and home goods focus

Augment leans into view-in-your-space for physical goods. Less emphasis on try-on, more on "place this couch in my living room." Solid pick for furniture, fixtures, and large appliances.

Auglio Eyewear Virtual Try-On

Auglio specializes in eyewear and jewelry with face-tracking. Sharper than generalist tools for glasses specifically. The Banuba team published a useful comparison of Shopify try-on apps that's worth skimming.

LEVAR — for merchants who want a managed service

LEVAR includes 3D model creation, hosting, and an interactive viewer under one subscription. Higher cost, but you offload the entire 3D pipeline.

App selection checklist

• Does it support your product category with native tracking (face, hand, body, or room)?
• Does it produce or accept GLB/USDZ, so you own the assets if you switch vendors?
• What's the monthly minimum, and is there per-model pricing on top?
• How does it impact product-page Core Web Vitals?
• Does it fail gracefully on older devices, or break the page?

Integrating Model-Viewer in Your Shopify Theme

If you're on Dawn, Horizon, or any modern Online Store 2.0 theme, 3D model media renders automatically — no code required. But once you start customizing (custom sections, landing pages, non-product contexts), you'll touch Liquid directly.

The one-liner: model_viewer_tag

Shopify exposes a model_viewer_tag Liquid filter that outputs a complete, configured <model-viewer> element. Drop this into any section where you want a product's 3D model to render:

{% for media in product.media %}
  {% if media.media_type == 'model' %}
    {{ media | model_viewer_tag }}
  {% endif %}
{% endfor %}

Overriding defaults

Pass attributes as filter arguments to customize the viewer:

{{ media | model_viewer_tag:
     image_size: '800x',
     auto_rotate: true,
     camera_controls: true,
     ar: true,
     ar_modes: 'webxr scene-viewer quick-look'
}}

The ar_modes string is load-bearing — it tells the component which AR backends to attempt, in order. For a detailed walkthrough of the component's behavior and the full attribute list, the Shopify Dawn theme's main-product.liquid is the canonical reference.

Custom themes and headless storefronts

If you're running a headless setup with Hydrogen or a custom React frontend, you'll import <model-viewer> as a web component directly from the Google model-viewer library. The API is identical to what Shopify renders server-side — same attributes, same events. Our guide to headless Shopify with Hydrogen covers the broader architectural pattern if you're evaluating the switch.

CTA placement

The default "View in your space" button sits below the 3D canvas. On mobile, many merchants see higher engagement when the AR CTA is surfaced as a secondary button near the add-to-cart — not buried in the media gallery. Test placement as part of your broader product page optimization strategy.

Performance: Keeping AR From Wrecking Your Page Speed

A 15 MB GLB loading on every product page is a Core Web Vitals disaster. Here's how to keep AR performant.

Lazy-load the model

Set loading="lazy" on the <model-viewer> element and reveal="interaction" so the model only fetches when the user clicks to view it. This moves the 3D payload off the critical render path entirely.

Optimize assets aggressively

• Draco compress every GLB (usually 60-80% size reduction)
• Basis Universal or KTX2 texture compression on all PBR maps
• Polygon budget: under 50K triangles for most products; 100K max for hero items
• Texture budget: 2K maps, not 4K
• Bake normal maps rather than relying on high-poly geometry for surface detail

Measure before and after

Run Lighthouse on your product page before adding AR. Run it again after. If your Largest Contentful Paint (LCP) jumps more than 200ms, your model is too heavy or it's not lazy-loading. Don't accept the regression — fix the asset.

The mobile reality check

Remember that roughly 68% of Shopify sessions happen on mobile. Your 3D model is loading on a mid-range Android with limited memory and a 4G connection. Optimize for that device, not your desktop with a gigabit fiber line.

The ROI Math: When AR Pays Off (and When It Doesn't)

Let's put numbers to the pitch. The conversion lift from AR isn't uniform — it varies sharply by product category and implementation quality.

Published conversion lift data

• Shopify's platform data: products with 3D/AR content see 94% higher conversion rates on average, per Shopify's ROI on AR research
• Rebecca Minkoff case study: customers are 65% more likely to place an order after interacting with a product in AR, and 27% more likely after viewing it in 3D alone
• Return rate reduction: AR can lower returns by up to 40% by aligning customer expectations with the physical product
• Category-specific lifts: home furnishings see 60-80% conversion improvement; fashion accessories see 30-45%
• Industry benchmarks: Envive's roundup of online shopping conversion lift data shows AR try-on consistently outperforms static imagery across fashion verticals

A realistic example

A Shopify store doing $400K/year with a 2% conversion rate runs 200,000 sessions. Add AR to top 50 SKUs, at $30/model via AI generation = $1,500 investment. Assume a conservative 25% conversion lift on the AR-enabled product pages, which make up 40% of sessions. That's roughly $40K in additional annual revenue from a one-time asset investment. Payback period: under two weeks.

That math flips for smaller stores and certain categories. Skipping AR is sometimes the smarter move — here are the four scenarios where the ROI math doesn't work.

When to skip AR: commodity products

If you sell USB cables, printer paper, or industrial fasteners, customers aren't evaluating fit, look, or feel. They're evaluating price and delivery speed. AR adds cost and load time with no conversion upside. Skip it.

When to skip AR: under $50K/year in revenue

Below that threshold, your conversion problem isn't visualization — it's traffic. Investing $2,000 in 3D models when you're getting 500 monthly sessions is optimizing the wrong end of the funnel. Put that budget into SEO, paid acquisition, or reducing cart abandonment first.

When to skip AR: well-understood categories or shaky tech

T-shirts, generic mugs, candles with standard shapes — customers know what these look like. A 3D model of a plain white tee doesn't tell them anything new. Invest in lifestyle photography instead. Similarly, if your theme is a legacy Online Store 1.0 build, your hosting can't cache efficiently, or your product pages already fail Core Web Vitals, adding 3D assets makes the problem worse. Fix the foundation first, then add AR.

Common Mistakes to Avoid

Measuring AR Impact and Keeping It Accessible

Adding AR without measuring impact wastes the investment, and building it without accessibility considerations alienates a meaningful slice of your audience. Handle both together.

Event tracking with GA4

Fire a custom event when a user clicks "View in AR" or rotates the 3D model. In GA4, tag these as engagement events and create an audience for AR interactors. Compare conversion rates of AR-interactors vs. non-interactors on the same product pages.

Shopify native analytics and A/B testing

The Analytics → Reports section surfaces per-product conversion data. Tag AR-enabled products with a metafield, then filter reports to compare AR-enabled vs. non-AR SKUs over a 30-day window. Use Shopify's native A/B testing tooling or an app like Intelligems to test AR CTA placement — above-the-fold vs. buried in the gallery, desktop vs. mobile. Differences are typically 10-25%, which is meaningful at scale.

Return rate tracking

AR's biggest hidden payoff is return reduction. Tag orders that included an AR-interacted product and compare 30-day return rates to non-AR orders. If you see even a 15% reduction, that often dwarfs the conversion lift in dollar terms — returns eat margin in ways that conversion dashboards don't show.

Accessibility baseline

AR-first product pages can unintentionally exclude users. Hit this baseline on every implementation:

• Alt text on every 3D model describing the product as thoroughly as a photograph would
• Fallback 2D imagery that renders when <model-viewer> fails to load or when a user has disabled WebGL
• Keyboard navigation for rotating and zooming the 3D view (most themes handle this, but verify)
• Reduced motion respect — auto-rotate should be off for users with prefers-reduced-motion: reduce
• Clear indicators that a "View in AR" button will open the camera — surprise camera access is hostile UX

The Shopify theme ecosystem has come a long way on accessibility, but AR is still the weakest link in most implementations. If you build for the customers who need accommodations, the whole page gets better for everyone.

Getting Started: Your 30-Day AR Rollout Plan

Rather than trying to AR-enable your entire catalog in week one, use this phased rollout.

Week 1 — Pick your pilot. Identify 5-10 hero products with the highest revenue or traffic. These are your AR test bed.

Week 2 — Generate assets. Scan or commission 3D models for each pilot product. Upload both GLB and USDZ to each product in Shopify admin.

Week 3 — Integrate and measure. Confirm the 3D viewer appears correctly on desktop, iOS, and Android. Set up GA4 events on AR interactions. Run a before/after Lighthouse pass to confirm performance is intact.

Week 4 — Analyze and scale. Compare conversion rates on AR-enabled vs. control pages. If you see the lift, expand to the next 50 products. If you don't, audit the asset quality and viewer placement before giving up.

Connect with other merchants running similar tests in the Talk Shop community — category-specific results vary enormously, and benchmarking against peers beats working in isolation.

The Takeaway

AR product try-on isn't experimental anymore. The tooling is mature, the conversion data is consistent across studies, and Shopify's native support means you don't need a developer on retainer to ship it. The merchants who win are the ones who treat AR as a product page upgrade — not a marketing gimmick — and measure the impact with the same rigor they apply to every other conversion test.

If you sell physical goods with meaningful visual variation — apparel, eyewear, jewelry, furniture, home décor, footwear, accessories — AR belongs on your 2026 roadmap. Start with ten hero products, ship GLB and USDZ on both, measure the lift, and scale from there. For more tactical optimization guides, browse our conversion optimization library and the broader Talk Shop blog.

What's the one product on your store right now where a 3D model would change how customers understand it? That's your pilot. Ship it this month.

About Talk Shop

The Talk Shop team — insights from our community of Shopify developers, merchants, and experts.