OpenAI has launched Shopping Research, a new feature inside ChatGPT designed to help users compare products and make informed buying decisions without checking multiple websites. Instead of searching across platforms manually, users can simply describe what they need – such as “Find the quietest cordless vacuum,” “Help me choose between these bikes,” or “Suggest a gift for a 4-year-old who loves art” – and ChatGPT produces a personalized buyer’s guide within minutes.
OpenAI says millions of people already rely on ChatGPT to understand products, compare specs, and decide between alternatives. The new Shopping Research feature enhances this behavior by offering:
- Dynamic product comparisons
- Tradeoff analysis based on preferences
- Updated pricing, availability, and reviews
- Category-specific guidance across electronics, home, beauty, outdoor, kitchen, sports, and more
For simple queries, ChatGPT still responds normally. When comparisons or tradeoffs are needed, Shopping Research generates structured, in-depth results.
How Shopping Research works
1. Start with a shopping query
Users can begin by typing a shopping-related question or selecting “shopping research” from the (+) menu. A guided interface appears, prompting for:
- Budget
- Preferences
- Type of user (adult, child, pet, etc.)
- Important features
If ChatGPT memory is enabled, previous conversations – like gaming, travel, or fitness preferences – can influence results.
2. Real-time product research
Behind the scenes, ChatGPT pulls updated information across the internet:
- Prices and availability
- Technical specifications
- Reviews
- Images and variants
Users can refine results with “Not interested” or “More like this” to adjust recommendations on the fly.
3. A personalized buyer’s guide
After analysis, ChatGPT generates a detailed shopping guide that includes:
- Product recommendations
- Key differences
- Pros and cons
- Tradeoffs
- Verified retailer links
OpenAI says Instant Checkout for supported merchants will arrive in a future update.
4. Deep integration with ChatGPT Pulse
For Pro users, ChatGPT Pulse can automatically surface guides. Example: After discussing an e-bike, Pulse might suggest accessories or maintenance tools.
Shopping Research is powered by a specialized GPT-5 mini model trained with reinforcement learning for shopping-oriented tasks.
Key capabilities include:
- Synthesizing multiple sources
- Accurate spec and feature matching
- Contextual preference understanding
- Real-time updates
- Cited sources and spam filtering
OpenAI notes that while accuracy is high, occasional errors in pricing or availability may occur, so users should verify before purchasing.
Privacy and Transparency
- User conversations are not shared with retailers
- Data comes from public retail sources
- Citations appear throughout the guide
- Low-quality and spam sources are filtered out
Shopping Research is rolling out now across the ChatGPT Free, Go, Plus, and Pro plans on mobile and web.
During the holiday season, usage will be nearly unlimited across supported plans.
