---
slug: vertex-ai
competitorName: Google Vertex AI
title: 'Google Vertex AI Virtual Try-On vs Genlook: Try-On API Compared'
description: >-
  Google's Vertex AI Virtual Try-On costs $0.06 per image and covers clothing.
  Genlook is a try-on API at $0.02 for any product type. Benchmark results
  inside.
keywords:
  - Vertex AI virtual try-on
  - Google virtual try-on API
  - virtual-try-on-001
  - Vertex AI try-on pricing
  - virtual try on API
eyebrow: Vertex AI try-on alternative
h1: Genlook vs. Google Vertex AI Virtual Try-On
tldr: >-
  Google's Virtual Try-On model renders clothing very well, at $0.06 an image
  inside a Google Cloud project. Genlook is a try-on API at $0.02 that covers
  every product type and handles the work around the model.
primaryCtaText: Get API keys
primaryCtaHref: 'https://platform.genlook.app'
secondaryCtaText: Read the docs →
secondaryCtaHref: /docs/tryon-api/introduction
author: thibault
lastChecked: '2026-10-07'
sources:
  - label: >-
      virtual-try-on-001 model page: launch stage, deprecation and retirement
      dates, specs, regions
    url: >-
      https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/vto/virtual-try-on-001
  - label: >-
      Generate virtual try-on images: supported models, REST and Python usage,
      free credits
    url: >-
      https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/capabilities/generate-virtual-try-on-images
  - label: 'Virtual try-on API parameters: sample count, safety settings, watermark'
    url: >-
      https://docs.cloud.google.com/vertex-ai/generative-ai/docs/reference/rest/Shared.Types/VirtualTryOnModelParams
  - label: 'Vertex AI generative AI pricing: $0.06 per image, savings plan prices'
    url: 'https://docs.cloud.google.com/vertex-ai/generative-ai/pricing'
  - label: 'Vertex AI release notes: preview, quality updates, general availability'
    url: 'https://docs.cloud.google.com/vertex-ai/generative-ai/docs/release-notes'
  - label: 'Vertex AI virtual-try-on-001 model page: request quota'
    url: >-
      https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/imagen/virtual-try-on-001
  - label: Genlook Try-On API pricing
    url: 'https://genlook.app/developers#pricing'
noindex: false
canonicalPath: /developers/vs/vertex-ai
---

<Verdict label="01 · The short verdict" title="A strong model, or a finished try-on API." description="Both produce excellent clothing try-ons. They differ in price, product coverage and how much you build around the call.">
  <VerdictColumn name="Google Vertex AI" bestFor="Built for Google Cloud teams">
    <Point type="pro">Excellent clothing results in our benchmark, close to Genlook's</Point>
    <Point type="pro">Runs in your Google Cloud project, EU regions included</Point>
    <Point type="pro">Up to 4 output images per request</Point>
    <Point type="con">$0.06 per image; $0.048 only on a 3-year savings plan</Point>
    <Point type="con">Clothing only, one product per request, synchronous calls</Point>
    <Point type="con">virtual-try-on-001 is deprecated; retirement set for March 2027</Point>
  </VerdictColumn>
  <VerdictColumn name="Genlook" bestFor="Built for shipping try-on">
    <Point type="pro">$0.02 per try-on on plans, down to $0.015</Point>
    <Point type="pro">Any product type: clothing, shoes, glasses, jewelry, hats, wigs</Point>
    <Point type="pro">Async with webhooks, plus batches at half a credit</Point>
    <Point type="pro">One stable contract while the engine improves behind it</Point>
  </VerdictColumn>
</Verdict>

<ComparisonTable label="02 · Feature by feature" title="Where each API stands." subtitle="Checked against Google Cloud's model pages, pricing and release notes on the date above.">
  <Row feature="Price per try-on" desc="What a generated image costs" genlook="$0.02 on plans, down to $0.015; $0.04 pay as you go" competitor="$0.06; $0.054 or $0.048 on 1- or 3-year savings plans" highlightGenlook />
  <Row feature="Free start" desc="Trying before paying" genlook="10 free credits, keys in minutes" competitor="$300 Google Cloud credit for new customers" />
  <Row feature="Speed" desc="Time to a result" genlook="8.2s median" competitor="No published figure; 11 to 12.5s in our benchmark" highlightGenlook />
  <Row feature="Product types" desc="What can be tried on" genlook="Any product type, one endpoint" competitor="Clothing; no published category list" highlightGenlook />
  <Row feature="Delivery" desc="How results come back" genlook="Sync call, or async with webhooks; batches at half a credit" competitor="Synchronous predict call; no batch inference" highlightGenlook />
  <Row feature="Model lifecycle" desc="What happens when models change" genlook="Same endpoint; engine updates behind it" competitor="Deprecated Sept 14, 2026; retirement March 15, 2027" highlightGenlook />
  <Row feature="Setup" desc="What you need first" genlook="An API key" competitor="Google Cloud project, billing, IAM and the Vertex AI API" highlightGenlook />
  <Row feature="End-user data" desc="Lifecycle beyond the generation" genlook="User records, deletion API, image auto-expiry" competitor="Not offered; storage and records are yours" highlightGenlook />
  <Row feature="Data use" desc="Training on your images" genlook="Never used to train models" competitor="Not used for training without your permission" equal />
</ComparisonTable>

<BenchmarkSection
  label="Same photos, same day"
  title="Genlook and Vertex AI, side by side."
  note="From our September 2026 benchmark: the same shopper photo and product photo sent to each API, default settings, one run each, unretouched. Times and prices are what each run took and cost; Genlook's time is its production median at the time."
  methodLabel="See all 10 APIs on 8 garments"
  methodHref="/blog/best-virtual-try-on-apis-2026"
>
  <BenchmarkRow
    set="trench-coat"
    competitor="vertex"
    garment="Trench coat, packshot"
    competitorLabel="Google Vertex AI"
    genlookStats="9.8s median · $0.020"
    competitorStats="12.5s · $0.060"
    verdict="Vertex dressed the coat but removed the shopper's jeans. Genlook kept them and matched the knee length."
  />
  <BenchmarkRow
    set="lace-gown"
    competitor="vertex"
    garment="Lace wedding gown"
    competitorLabel="Google Vertex AI"
    genlookStats="9.8s median · $0.020"
    competitorStats="11.6s · $0.060"
    verdict="Both render the lace bodice and tulle skirt faithfully. A close call."
  />
  <BenchmarkRow
    set="football-shirt"
    competitor="vertex"
    garment="Football shirt with sponsor text"
    competitorLabel="Google Vertex AI"
    genlookStats="9.8s median · $0.020"
    competitorStats="12.0s · $0.060"
    verdict="Both keep the sponsor text readable and the stripes aligned."
  />
</BenchmarkSection>

<ProseSection label="03 · The real difference" title="A model endpoint, or the whole try-on layer">
On clothing, Google's model is one of the best we have tested. In our September benchmark it matched Genlook on most garments, and its only unwanted change in eight tries was removing a shopper's jeans under a long coat. If your stack already lives in Google Cloud and you only need garments, it is a serious option.

What you get is a model endpoint. Each call takes one person image and one product image and returns up to four images synchronously; there are no webhooks, no batch mode and no records of your end users. Coverage is clothing, with no published category list for shoes, glasses or jewelry. And the model has its own lifecycle: virtual-try-on-001 reached general availability in January 2026, was marked deprecated on September 14, 2026, and is set to retire in March 2027, with Google's docs now listing Gemini image models as try-on options. Each change is a migration on your side.

### How Genlook handles it

Genlook sells the try-on layer as one API. The same endpoint covers clothing, shoes, glasses, jewelry, hats and wigs, answers synchronously or through webhooks, and runs large jobs as batches at half a credit. End users are records you can delete on request, uploaded images expire on a schedule you choose, and when a better model appears the engine changes behind the same contract, listed in the [changelog](/docs/tryon-api/changelog). A try-on costs $0.02 on a plan and $0.015 at volume, a third of Vertex's list price.

If you need clothing only, run everything inside Google Cloud and want to own the integration, Vertex is a strong pick. If you want every product type, a stable contract and the operations included, start with Genlook.
</ProseSection>

<FeatureGrid label="04 · In practice" title="What the Genlook API is tuned for.">
  <Feature title="Any product type" description="Clothing, shoes, glasses, jewelry, hats and wigs through one endpoint, with no category parameter to manage." />
  <Feature title="A contract that stays put" description="Engine improvements land behind the same endpoint. No model ids to track, no deprecation notices to plan around." />
  <Feature title="Async when you need it" description="Webhooks for completed and failed generations, and batches of up to 500 try-ons at half a credit each." />
  <Feature title="End-user data handled" description="User records, a deletion endpoint for privacy requests, and image retention of 1, 3 or 7 days." />
</FeatureGrid>

<Steps label="05 · Switching" title="Moving from Vertex AI.">
  <Step number="1" title="Create a key" description="Self-serve at platform.genlook.app, 10 free credits included. No cloud project, billing account or IAM setup." />
  <Step number="2" title="Replace the predict call" description="Send the same person and product images to Genlook's try-on endpoint, synchronously or with a webhook. Shoes, glasses and jewelry work through the same call." />
  <Step number="3" title="Go live" description="Plans start at $20 a month for 1,000 try-ons ($0.02 each), down to $0.015 at volume, or pay as you go at $0.04. No model deprecation dates to track." />
</Steps>

<Faq label="06 · FAQ" title="Questions, answered.">
  <FaqItem question="How much does Google's Virtual Try-On cost?">
$0.06 per generated image on Vertex AI. Google's savings plans bring it to $0.054 with a 1-year commitment or $0.048 with a 3-year commitment. Genlook costs $0.02 per try-on on a plan, down to $0.015 at volume, or $0.04 pay as you go.
  </FaqItem>
  <FaqItem question="Is Vertex AI's try-on better than Genlook's?">
On clothing, the two were close in our benchmark: both rendered lace, prints and sponsor text faithfully. Vertex removed a shopper's jeans under a long coat once; Genlook made no unwanted changes in that run. The bigger differences are coverage (Vertex is clothing only), price and the integration work around the call.
  </FaqItem>
  <FaqItem question="Is virtual-try-on-001 being retired?">
Google's model page lists a deprecation date of September 14, 2026 and a retirement date of March 15, 2027, and its try-on guide now also lists Gemini image models. Plan a migration if you build on it. Genlook keeps one endpoint and updates the engine behind it.
  </FaqItem>
  <FaqItem question="Does Vertex AI try-on work for shoes, glasses or jewelry?">
Google's docs describe it for clothing and publish no category list for accessories. Genlook covers clothing, shoes, glasses, jewelry, hats and wigs through the same endpoint.
  </FaqItem>
  <FaqItem question="Can I test both?">
Yes. New Google Cloud customers get $300 in credit, and Genlook accounts start with 10 free credits. Send the same photos to both and compare the results on your own catalog.
  </FaqItem>
</Faq>

<Sources />

<CtaBlock
  title="Build try-on into your product."
  description="Self-serve keys, ten free credits, and two API calls to your first generation."
  primaryText="Get API keys"
  primaryHref="https://platform.genlook.app"
  secondaryText="Read the docs"
  secondaryHref="/docs/tryon-api/introduction"
/>