AI Commerce Search

What it does

AI Commerce Search (formerly Vertex AI Search for Commerce) takes in a retailer's product catalog and customer behavior events, then returns ranked search results, browse results and product recommendations through API calls. Its commerce-tuned language models interpret queries, and it can enrich catalog data with attributes and taxonomy using Google Shopping crawlers and vision models.

Retail teams use it to replace or upgrade on-site search, tune rankings against goals such as conversion rate or revenue per session, and run multi-turn conversations that help shoppers narrow broad queries. A Merchandising console lets business users boost or bury results, launch promotions and reroute results when items are out of stock. Recommendations can also feed mobile apps, email campaigns, in-store kiosks and call centers.

It suits e-commerce and retail organizations with engineering resources, or an implementation partner, to integrate an API into their storefront. It runs on Google Cloud, so a Google Cloud account and billing are needed, and a sales contact route is offered. Charges are usage-based, so costs scale with traffic, and recommendations also bill for model training and tuning. The service is a back-end capability rather than a standalone shopper-facing app.

What it can do

• Personalized product search and browse results
• Product recommendations across web, mobile, email, kiosks and call centers
• Conversational Commerce agent for multi-turn query refinement
• Merchandising console to boost, bury and reroute results
• Catalog enrichment with attributes and taxonomy

Example workflows

Upgrade storefront search

Inputs: product catalog and user event data. Steps: ingest catalog and events, connect storefront search to the Search method, set ranking goals. Output: personalized search and browse results for shoppers.

Run a promotion

Inputs: a campaign plan and product list. Steps: open the Merchandising console, boost promoted items, reroute results for out-of-stock products, test the changes. Output: adjusted result ordering without code changes.

Before you choose

• Search is billed per 1,000 queries, so high-traffic sites should model costs.
• Recommendations add node-hour charges for training and tuning on top of prediction requests.
• Deep integration is required; it is an API service, not a ready-made storefront widget.

Listed under

• Marketing
• Ecommerce
• Conversational commerce
• E-commerce
• Enterprise
• Merchandising
• Product recommendations

Get started

Google Cloud console also available for building

1. Create or sign in to a Google Cloud account and billing account
2. Ingest your product catalog and user events
3. Integrate Search or recommendation calls into your site or app, or work with a partner

Recommendations: $0.27/$0.18/$0.10 per 1,000 predictions by monthly tier, plus $2.50 per node-hour for training and tuning. $600 trial credits for recommendations expire after six months. Custom quotes available via sales.

Source: cloud.google.com/retail/pricing, checked 28 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 28 Oct 2026. Each source supports part of the overview; availability and prices can change.

• AI Commerce Search | Google Cloud ↗cloud.google.com
• AI Commerce Search pricing | Google Cloud ↗cloud.google.com

Company website

cloud.google.com: product, pricing and policy pages

YC directory mirror

Batch and company name

Web search

Confirms the official site

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

Last checked 28 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.