DATED EXAMPLE
Which sources appear for AI comparison questions?
What does an AI web-search run look for when comparing Ahrefs and Semrush?
The run split the question into a broad comparison, one official-pricing query per product and a small-business review query. OpenAI returned the sources on the same search action, so that action-level connection can be shown truthfully.
01
Fixed protocol
- Provider
- OpenAI API
- Model
- GPT-5.6 Luna
- Tool
- web_search · low context
- Limits
- 8 searches · 500 output tokens
- Provider storage
- store: false
- Version
- native-fanout-tool/1.0.0
02
Observed queries
Original query strings from the API response. These are not post-run search ideas.
- Ahrefs vs Semrush small business pricing features 2026
- Ahrefs pricing official
- Semrush pricing official
- Ahrefs Semrush review small business SEO comparison
Source domains on the search action
- ahrefs.com
- semrush.com
- help.ahrefs.com
- stackcompare.net
- formativedigital.com
- insidepro360.com
- softoraguide.com
- angletutoring.com
- static.semrush.com
- sec.gov
- youtube.com
- reddit.com
- cdnc.heyzine.com
- Input
- 12,876 Tokens
- Output
- 475 Tokens
- Estimate
- $0.0131
03
What this supports
- Four query strings were exposed.
- Two queries explicitly asked for official pricing.
- The action returned first-party product domains, reviews, Reddit, YouTube and an SEC domain.
04
What it does not prove
- The API linked sources to the search action, not to one exact query inside that four-query action.
- A citation is not an endorsement or ranking.
This example is not an independent model benchmark. It documents individual API observations under a published protocol and does not reproduce the ChatGPT or Gemini consumer interface.