From ChatGPT to Naver,
AI search answer collection API
Collect the answers and cited sources people see in AI services.
Get consistent JSON results across services for analysis and monitoring.
Reduce repetitive integration and response formatting across AI services.
Reduce per-service authentication and response formatting.Specify which services to collect with Chatto API and receive consistent answers and sources.
When integrating directly
- Authenticate each service
- Convert request formats by service
- Extract answers from responses
- Normalize and store the results
- Check and retry failed requests
- Update code when services change
With Chatto API
AI answers: from collection requests to results.
Submit a question and the AI services to collect.Check status using the run ID, then retrieve answers and sources as JSON.Expanded search queries are available only for supported services.
1curl -X POST https://platform-api.trychainshift.ai/api/v1beta/platform/runs \2 -H "API-Key: cs_live_9a2f..." \3 -H "Content-Type: application/json" \4 -d '{5 "models": [6 "CHATGPT", "GEMINI", "CLAUDE", "PERPLEXITY",7 "GOOGLE_AI_OVERVIEW", "GOOGLE_AI_MODE",8 "NAVER_AI_BRIEFING", "NAVER_AI_TAB"9 ],10 "promptIDs": ["n9Ixmd1VD1pOTfVjppBJjQ"],11 "region": "KR",12 "repeatCount": 413 }'1{2 "status": "SUCCESS",3 // Expanded queries used by AI to produce the answer4 "fanouts": [5 "Measuring brand visibility in AI answers",6 "Recommended brand mention tracking tools",7 "Monitoring AI search brand mentions"8 ],9 // Answer content for the prompt10 "content": "To see how often a brand appears in AI answers,11 1. Chatto — collect answers from multiple AI services…",12 // Sources for the answer13 "sources": [14 { "title": "Chatto documentation", "url": "..." },15 { "title": "AI search tools comparison blog", "url": "..." }16 ],17 ...18}- ChatGPT
- Gemini
- Claude
- Perplexity
- Google AI Overview
- Google AI Mode
- Naver AI Briefing
- Naver AI Tab
AI answers and cited sources in one consistent JSON format.
Collect answers and sources from real AI service interfaces.Set countries, retrieve results, and check failed runs through the API.
Integration
Use one authentication and request setup across AI services.
- ChatG…
- Gemini
- Claude
- Perpl…
- Googl…
- Googl…
Response format
Answers and sources from every service use the same field structure.
Answers from real user environments
Collect the answers visible in the AI services people actually use.
Retrieve collection results
Review answers with their collection times and compare previous results.
08-05 09:10
08-08 14:20
08-11 11:05
Answers by country
Choose a country to see the answers shown to users in that market.
South Korea Answer
Failure handling
Failed collection runs do not consume credits. Submit a new run for the same question if you need to collect it again.
Choose questions using search keywords, then analyze the answers.
Use search volumes and suggested or related keywords to choose questions.Analyze repeated expressions, cited sources, and expanded search queries in the answers.
Search keyword tools
BetaCheck search volumes and suggested or related keywords from Naver and Google search data.
- Search volume and keywords
- Get search volumes and suggested or related keywords
- Site analysis
- Extract key terms from a website
Use cases
- Check search demand
- Brand analysis
- Prompt design
Answer analysis tools
BetaAnalyze AI answers at scale to find repeated expressions, key sources, and leading fanout queries.
- Expanded search queries
- Top fanout queries
- Cited sources
- Top cited domains and content, with citation counts by page
Use cases
- Content citation status
- Brand associations
- Content writing
Compare API prices and estimate your collection costs.
Compare per-request prices by AI service.Estimate credits and payment based on your questions and collection runs.
01·Models
02·Volume
03·Estimate
- Selected models
- 4
- Estimated requests
- 800 times
- Estimated creditsLaunch offer 50% off
- 1,200600
- Estimated payment
- ₩12,000
Get an API key and start collecting AI answers.
Get an API key and use the request examples in our developer documentation.Specify your questions and AI services, then review the results.
AI answer collection API: frequently asked questions
Learn about collection methods, supported services, result data, and credit usage.
Collection methods
We run questions in the AI service environments people use, for your selected models and countries, and collect the answers shown on screen.
Official APIs return responses generated by developer-facing models.Chatto collects the answers users see when asking questions in services such as ChatGPT or Perplexity.
Generative AI can answer the same question differently over time.Repeated collection reveals recurring wording, sources, and trends beyond a single result.
Collection is asynchronous.Check status using the ID returned by your run request, then retrieve answers once the run is complete.
Supported services
Eight models are supported: ChatGPT, Gemini, Claude, Perplexity, Google AI Overview, Google AI Mode, Naver AI Briefing, and Naver AI Tab.
Answers from different models use one JSON schema, so you can store or analyze them without separate formatting for each model.
Retrieve the answer text and source URLs, titles, and domains.Expanded search queries (fanout) are available for ChatGPT, Perplexity, and Claude; other services return an empty array.Sources and expanded queries can be empty for individual answers.
Choose South Korea, the United States, the United Kingdom, Germany, Japan, Canada, Australia, Thailand, Vietnam, or Singapore.Naver AI Briefing and AI Tab support South Korea (KR) only; the other AI services support all ten regions.
Pricing and operations
Collection runs equal questions × AI services × repetitions.Credits are deducted at the applicable rate for each completed run; failed runs are not charged.Run requests, status checks, and answer retrieval are not all counted as the same billing unit.See the calculator above for current rates and discounts.
Credit purchase prices stay the same.During the launch offer, collection credits are automatically reduced from 1.5 to 0.75 per request.
No.Failed answer collection requests do not consume credits.
You may process collected data into reports, dashboards, and consulting deliverables for clients.Public use requires a “Powered by Chatto” credit.Reselling the API or redistributing raw data in bulk requires prior agreement.
Yes.Contact us for large-scale collection or integration with your existing data pipeline.We will advise based on your collection volume and use case.