Fan-Out Queries

GET /api/v1/fanout-queries returns the fan-out queries recorded for your tracked prompts: the searches an engine ran while producing an answer. Each row belongs to one run, so you can join back to runs and prompts with run_id and prompt_id.

Filters

Standard parameters (limit, offset, format=csv) work as described in Pagination, filtering, and errors.

Key fields

• run_id: the run this query was recorded for
• prompt_id: the tracked prompt behind the run
• engine_id: the engine that produced the answer (ids come from GET /api/v1/engines)
• query_index: position of the query within the run's fan-out (0-based)
• query_text: the query itself
• query_type: search, shopping, or synthetic
• source: observed for provider-reported queries; today all rows are observed, and modeled is reserved so clients will not need a breaking change if modeled rows are added later
• mentioned_brand_ids_json: tracked brands named inside the query text, joinable to brands
• captured_at: when the fan-out was recorded

Rows are ordered by captured_at descending with run_id and query_index as tiebreaks, so offset pagination is stable.

Example

curl -H "Authorization: Bearer $AEO_API_KEY" \
  "https://app.citlyze.com/api/v1/fanout-queries?source=observed&since=2026-08-01"

Use this resource to see which sub-queries decide the answers you track, which of them mention your brand or competitors, and how fan-out differs between engines.