OpenAI Launches Virtual Try-On for ChatGPT Shopping

Overview

OpenAI has expanded the commercial capabilities of its flagship assistant with the global rollout of two new shopping features. The updates introduce a virtual try-on tool and a dedicated Favorites system, aiming to bridge the gap between conversational AI and e-commerce.

The launch positions ChatGPT as a more direct participant in the online retail journey. By leveraging its recently updated image generation model, the company is attempting to offer a visual and interactive shopping experience that rivals dedicated search engines and social platforms.

New Features and Capabilities

The core innovation is the virtual try-on function, which allows users to visualize how clothing and accessories might appear on their own bodies. Users can upload a selfie or a full-body photograph to generate realistic overlays of the desired item.

This feature integrates directly into ChatGPT’s shopping results. When users encounter products during a search, a new Try On button becomes available. Clicking it initiates the generation process, using the user’s provided photo as the canvas.

OpenAI stated that the updates leverage the newly launched ChatGPT Images 2.5 model, which reportedly produces more natural lighting and richer textures while reducing generation latency.

Beyond direct try-ons, the tool supports a unique reverse-image workflow. Users can upload a screenshot of an item from a website or a photo of a celebrity outfit. They can then instruct ChatGPT to virtually place that item on themselves, effectively bridging the gap between inspiration and personal visualization.

The Favorites System

Complementing the visual tools is a new Favorites function designed to streamline product discovery and retention. This feature allows users to save items they discover during their conversations to a dedicated library within the app.

The library serves as a centralized repository for future reference. Notably, OpenAI has integrated the try-on images into this same library, ensuring that visual experiments and product saves are stored in one cohesive location.

This consolidation addresses a common pain point in digital shopping, where users often lose track of items they wish to purchase at a later date. By keeping both the saved products and the visual simulations together, OpenAI hopes to create a more frictionless path to conversion.

Advanced Shopping Assistants

OpenAI is also encouraging users to adopt a more descriptive approach to shopping. Instead of searching for specific keywords, users can describe a style or aesthetic they wish to achieve. ChatGPT can then curate a list of pieces needed to complete that look.

Furthermore, the assistant can identify items worn by celebrities in uploaded photos. If a user shares an image of a public figure, they can ask ChatGPT to find similar products available for purchase. This capability mirrors features long established by Pinterest and Google, signaling OpenAI’s intent to compete in the visual discovery space.

Strategic Context and Competition

The entry into virtual try-on comes at a time when major tech giants are already entrenched in this domain. Google launched its own virtual try-on capabilities last year, giving it a significant head start in user familiarity and market penetration.

OpenAI’s move follows a period of experimentation with various consumer-facing features. The company previously pivoted away from an instant checkout feature, which did not perform as expected. This history suggests that OpenAI is carefully calibrating its e-commerce ambitions to avoid overreach while still capturing valuable user engagement.

The broader landscape of AI-driven shopping is also seeing increased activity from agentic startups. Competitors like Instinct have begun pushing proactive product recommendations to users. However, these efforts have faced criticism, with some observers arguing that such recommendations blur the line between helpful assistance and intrusive advertising.

The updates arrive as AI assistants explore consumer use cases around shopping, though OpenAI’s previous instant checkout idea was dropped due to poor performance.

By focusing on visualization and curation rather than direct transaction completion, OpenAI may be attempting to sidestep the trust issues that plagued its earlier checkout experiments. The strategy appears to be about enhancing the discovery phase of shopping rather than replacing the retailer’s checkout flow entirely.

User Experience and Implications

For the average consumer, these features represent a shift toward more conversational and visual search experiences. The ability to see how an item looks on one’s own body without physically trying it on could reduce return rates and increase confidence in online purchases.

The integration of image generation into shopping results also highlights OpenAI’s continued investment in its ChatGPT Images 2.5 infrastructure. Improvements in lighting, texture, and speed are likely aimed at making these generated images indistinguishable from traditional product photography.

However, the success of these features remains uncertain. While the technology is impressive, competing against established platforms like Google and Pinterest requires not just novelty but habitual usage. Users must decide whether ChatGPT is the right starting point for their shopping journeys.

As the holiday shopping season approaches, the impact of these tools will become clearer. For now, OpenAI has made a decisive move to position ChatGPT as a comprehensive assistant for not just information retrieval, but for tangible consumer decisions.