Monitor ChatGPT Responses
ChatGPT
Extract ChatGPT
Extract structured data from ChatGPT — answer text, cited sources, shopping cards, brand entities, and ads — for any prompt across supported regions
POST
Monitor ChatGPT Responses
Overview
The ChatGPT endpoint extracts structured data from ChatGPT with features including shopping cards, brand entities, map entries, raw response data, and query fan-out. Useful for e-commerce tracking, local business monitoring, and detailed analysis.Web search enabledThis endpoint automatically enables ChatGPT’s web search mode for all requests, so responses include current information from the web with source citations.
Unique features
- Sources: Sources include publication dates and footnote indicators for distinguishing primary from secondary references
- Shopping cards: Extracts structured product information with pricing, ratings, offers, and commercial details
- Inline products: Individual product references with pricing, offers, and rendering hints for inline display
- Entity extraction: Identifies and extracts named entities like products, brands, and concepts mentioned in responses
- Map entries: Extracts business and place information including ratings, reviews, contact details, and location data
- Citation pills: Captures inline citations with metadata including labels, descriptions, domains, and publication dates
- Ads: Extracts structured ad information including advertiser branding and product carousel cards
- Raw response access: Includes the full streaming response payload for debugging
- Query fan-out insights: Reveals the query fan-out ChatGPT used to generate responses
Request parameters
Uses common parameters. All ChatGPT-specificinclude.* flags are listed in the auto-generated request schema below.
Additional credit costEnabling any of
include.rawResponse, include.searchQueries, include.ads, or include.shopping (or any combination) adds +2 credits to the base cost.Response objects
The response includes the following sections. See each subpage for the full schema and examples.Response schema
Includes common response fields plus:Common questions
How does cloro retrieve ChatGPT responses?
cloro does not call the ChatGPT API. Each request runs through a real browser session against ChatGPT’s web interface, routed through a proxy in the country (and, where supported, state) you specify, with web search enabled. The structured fields underresult.* are extracted from that page, and the raw streaming SSE events are available via include.rawResponse: true. See How does cloro retrieve data from AI providers? for the full explanation and implications (model routing, auth walls, geo behaviour).
Can I get the search queries that ChatGPT uses?
Yes, you can get the actual search queries (fan-out queries) that ChatGPT uses internally by settinginclude.searchQueries: true in your ChatGPT requests. The same flag also returns result.mapSearchQueries: the queries ChatGPT sends to the maps tool to build the map block, which are separate from the web-search fan-out in result.searchQueries.
Are query fan-outs available for all ChatGPT responses?
ChatGPT usesgpt-5-3 and gpt-5-3-mini that include fan-out queries; newer models may not include explicit fan-outs. ChatGPT automatically selects which model to use for each request, so we cannot control the presence of query fan-outs.
Are query fan-outs available for other providers?
For a complete comparison of query fan-out support across all providers, see the Providers & pricing guide.Why is a fan-out query a single long string instead of separated keywords?
This is expected behaviour. ChatGPT’s search model decides the shape of each fan-out query, and it often emits a single long natural-language query rather than a comma-separated keyword list. The length and structure vary between runs, even for the same prompt, and cloro returns the queries exactly as ChatGPT generates them without reformatting. Community write-ups such as this LinkedIn post by Stefan Landwehr document the same behaviour.Why aren’t shopping cards appearing in my responses?
Shopping cards only appear when the prompt is related to products or shopping:Which providers support shopping cards?
For a complete comparison of shopping card support across all providers, see the Providers & pricing guide.My non-English prompts are returning English fan-out queries. Is that expected?
Yes. When you submit a non-English prompt, ChatGPT may still generate some or all of its internal search queries (result.searchQueries) in English. A roughly 50/50 English/local-language split has been observed for non-English prompts. This is upstream ChatGPT behavior — cloro returns the queries exactly as ChatGPT generates them. If you’re using fan-out for multilingual GEO monitoring, filter or group by query language as needed.
What’s the difference between shopping cards and inline products?
Shopping cards are grouped collections of products displayed together (like a carousel):- Contain multiple products
- Include category tags
- For browsing multiple options
- Single product per entry
- Include pricing, offers, images, and ratings
- Can be embedded inline in text, comparison tables, or featured displays
- Have rendering hints (inline, hero, block)
include.shopping: true is set in the request.Authorizations
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Body
application/json
Request parameters for monitoring ChatGPT responses
The prompt to send to ChatGPT
Required string length:
1 - 10000Example:
"What do you know about Acme Corp?"
Country/region code for localized response
Example:
"US"
Optional flags for including additional response formats
Example:
State code for sub-country geo-targeting (e.g., "CA"). Only valid with country "US".
Required string length:
2Pattern:
^[A-Z]{2}$Example:
"CA"