# Glance

> Glance AI is an intelligent shopping agent from Glance AI, Inc. that helps people discover and buy fashion and other products through personalised recommendations. A crew of AI agents curates a shopping feed that learns each user's style, predicts trends and lets users buy looks directly from official brand stores. Glance AI is available on phones and TVs, and the company was founded in 2019 and is part of the InMobi Group.

## Product & help

- [Glance AI home](https://glance.com/us): Overview of the Glance AI shopping agent and how it works.
- [Help centre](https://glance.com/us/glanceai/help): Help for Glance AI, covering shopping, your account and your feed.
- [Getting started](https://glance.com/us/glanceai/help/getting-started): How to set up Glance AI and start shopping.
- [Using Glance AI](https://glance.com/us/glanceai/help/using-glance-ai): How to use the shopping feed, the AI assistant and brand stores.
- [Request a demo](https://glance.com/us/glance-commerce-demo-request): Request a demo of Glance AI's commerce platform.
- [Contact us](https://glance.com/us/help/contact-us): How to get in touch with the Glance AI support team.
- [Download the app](https://glance.com/us): Get Glance AI on Google Play, the App Store or the Samsung Galaxy Store.

## Company

- [About us](https://glance.com/us/about-us): Who we are, our mission and our leadership team.
- [Business](https://glance.com/us/business): Partnership and business opportunities with Glance AI.
- [Advertising](https://glance.com/us/advertising): Advertising solutions for brands reaching Glance AI shoppers.
- [Partners](https://glance.com/partners): Brand, retail and technology partners.
- [Newsroom](https://glance.com/us/newsroom): Press releases and company announcements.
- [Careers](https://glance.com/us/careers): Open roles and how to join Glance AI.

## Trust & policies

- [Trust center](https://glance.com/us/trust-center): How Glance AI approaches trust, safety and user protection.
- [Privacy policy](https://glance.com/us/legal/privacy-policy-web-en): How Glance AI collects, uses and protects personal data.
- [Cookie policy](https://glance.com/us/legal/cookie-policy): How Glance AI uses cookies and similar technologies.
- [Security](https://glance.com/us/legal/security): Security practices and how to report a concern.
- [Compliance](https://glance.com/us/compliance): Compliance information for Glance AI.
- [End-user license agreement](https://glance.com/us/legal/eula-web-en): Terms for using the Glance AI app.
- [Advertiser terms](https://glance.com/us/legal/advertiser-terms): Terms that apply to advertisers using Glance AI.

## Top blog articles

- [Hey Siri, let Glance](https://glance.com/us/blogs/glanceai/shopping/hey-siri-lets-glance): How voice assistants and Glance AI work together for shopping discovery.
- [The future of agentic AI in shopping](https://glance.com/us/blogs/glanceai/shopping/future-of-agentic-ai-in-shopping): How AI agents may change the way people shop.
- [Agentic AI and autonomous decision-making](https://glance.com/us/blogs/glanceai/trends/agentic-ai-autonomous-decision-making): What autonomous AI means for everyday decisions.
- [Benefits of AI in e-commerce fashion](https://glance.com/us/blogs/glanceai/tech/benefits-of-ai-in-ecommerce-fashion): How AI is changing fashion shopping.
- [Glance AI blog](https://glance.com/us/blogs/glanceai): All articles on AI, fashion, shopping and technology.

## Developers

GlanceAI Fashion MCP Server: fashion product discovery, virtual try-on, AI outfit pairing and wishlist management via the Glance catalogue. Exposed over MCP (Streamable HTTP), Google A2A and AG-UI. Connect an MCP client to https://glance.com/mcp.

Discovery tools are public. Try-on, selfie, image and wishlist tools require OAuth 2.0 (PKCE).

### Endpoints

- [MCP Streamable HTTP](https://glance.com/mcp): Primary JSON-RPC endpoint (POST + GET), protocol version 2025-06-18
- [A2A](https://glance.com/a2a): Google Agent2Agent endpoint; agent card at https://glance.com/.well-known/agent.json
- [AG-UI SSE](https://glance.com/agui): CopilotKit / AG-UI server-sent event stream

### Tools: Discovery (public, no auth)

- search_fashion_products: Natural-language fashion search (e.g. "red summer dress under 2000"). Returns a product carousel.
- get_product_details: Full product detail by SKU: price, brand, images, availability.
- get_mix_and_match: Suggest complementary items to build a full outfit around a product.
- show_products_carousel: Render a set of products as a carousel artifact.

### Tools: Virtual try-on (auth and user selfie required)

- try_on_product: Render the user wearing a single product (by SKU) or a full outfit or collection (by collection name).
- get_user_looks: Retrieve the user's saved AI-generated looks.

### Tools: Selfie and image (auth required)

- request_user_selfie: Prompt the user to upload a selfie for try-on.
- check_selfie: Check whether the user already has a usable selfie.
- request_context_image: Present an image upload widget and return a ref_id.
- get_selfie: Fetch the user's stored selfie.

### Tools: Wishlist (auth required)

- show_wishlist: Render the user's current wishlist.

### Auth

- [OAuth authorization server metadata](https://glance.com/.well-known/oauth-authorization-server): OAuth 2.0 + PKCE discovery (RFC 8414)
- [OAuth protected resource metadata](https://glance.com/.well-known/oauth-protected-resource): Bearer token configuration for the MCP endpoint