Introducing Prompt Research

Prompt Research turns one topic into the questions people ask AI assistants about it, ranked by how many people search for that need each month.

What you get

• prompts: up to 50 questions written the way people ask an AI assistant, highest demand first, each with its monthly search volume and its share of the topic’s demand.
• The topic, persona, intent and funnel stage of every prompt, so you can group them by sub-field and buying stage.
• monitorSet: the prompts a GEO monitor should track, in the order to add them.
• brands: the competitor brands that draw search demand in the same space.

How it works

Send a seed topic such as “sunscreen” and a market. Prompt Research measures the monthly search demand behind the topic in that market, from the last 30 days in KR and the average month of the past 12 months in US, finds the needs people search for most and writes each one as the question a person would put to an AI assistant. KR returns Korean prompts and US returns English ones, and most runs finish in 30 to 60 seconds.

Add your brand and a short description of what it sells, and every prompt gets a fit score. Prompts that contain your brand names are left out, so a monitor built from them measures the mentions your brand earns on its own.

The Prompt research tab in the Playground and the new monitor form run the same research.

Send a request

Send the request with your API key in the Authorization header.

curl -X POST https://api.querying.ai/v1/prompt-research \
  -H "Authorization: Bearer $QUERYING_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "seed": "wireless earbuds", "country": "US", "limit": 20 }'

Request fields

seedstringRequired

The topic: 1 to 30 characters of letters, digits and single spaces, such as sunscreen.

countrystringRequired

KR or US. Prompts come in the market’s language.

limitinteger

Prompts to return, 1 to 50. Default 20.

excludearray

Up to 20 terms. A prompt containing one leaves the result.

monitorSizeinteger

Prompts in monitorSet, 1 to 50. Default 20.

monitorEnginesarray

The engines your monitor will run. Keeps monitorSize within a monitor’s 200 prompt × engine tasks.

brandstring

Your brand. Prompts naming it leave the result, so every prompt measures real demand.

brandDescriptionstring

Up to 500 characters on what your brand sells and to whom. Adds fit to each prompt and favors the prompts you answer.

webhook.urlstring

An HTTPS address on your server. The finished task is posted there the moment it completes.

Response

A finished task looks like this, trimmed for length.

{
  "success": true,
  "task": {
    "id": "ba954bab-33ec-4bb0-b175-a9b85258a4ac",
    "taskType": "PROMPT_RESEARCH",
    "status": "COMPLETED",
    "createdAt": "2026-10-02T15:55:29.285Z"
  },
  "credits": { "creditsToCharge": 12, "creditsCharged": 12 },
  "response": {
    "seed": "wireless earbuds",
    "country": "US",
    "language": "en",
    "seedMonthlySearchVolume": 165000,
    "promptsFound": 28,
    "prompts": [
      {
        "prompt": "Which noise-cancelling earbuds should I choose?",
        "topic": "Earbud features",
        "persona": null,
        "intent": "commercial",
        "funnelStage": "consideration",
        "brandMentions": "likely",
        "monthlySearchVolume": 118380,
        "demandShare": 0.2585
      },
      {
        "prompt": "What are the best wireless earbuds to buy?",
        "topic": "Choosing earbuds",
        "persona": null,
        "intent": "commercial",
        "funnelStage": "consideration",
        "brandMentions": "likely",
        "monthlySearchVolume": 78840,
        "demandShare": 0.1722
      }
    ],
    "monitorSet": {
      "prompts": [
        {
          "rank": 1,
          "prompt": "Which noise-cancelling earbuds should I choose?",
          "topic": "Earbud features",
          "persona": null,
          "intent": "commercial",
          "funnelStage": "consideration",
          "brandMentions": "likely",
          "monthlySearchVolume": 118380,
          "demandShare": 0.2585
        }
      ],
      "coverage": {
        "demandShare": 0.8406,
        "topics": { "total": 5, "covered": 5 },
        "personas": { "total": 9, "covered": 8 }
      }
    },
    "topics": [
      {
        "topic": "Choosing earbuds",
        "monthlySearchVolume": 191440,
        "promptCount": 9,
        "demandShare": 0.4181
      },
      {
        "topic": "Earbud features",
        "monthlySearchVolume": 161940,
        "promptCount": 6,
        "demandShare": 0.3537
      }
    ],
    "personas": [
      {
        "persona": "people who sleep with earbuds",
        "monthlySearchVolume": 17700,
        "promptCount": 1,
        "demandShare": 0.0387
      }
    ],
    "brands": [
      { "brand": "Beats", "monthlySearchVolume": 560980 },
      { "brand": "Apple", "monthlySearchVolume": 509920 },
      { "brand": "Samsung", "monthlySearchVolume": 91750 }
    ]
  }
}

task.statusstring

QUEUED, PROCESSING, then COMPLETED with the answer or FAILED with error.

credits.creditsChargedinteger

Credits charged for the finished task. A failed task charges 0.

response.promptsarray

Questions people ask AI assistants, highest demand first, with topic, persona, intent, funnelStage, monthlySearchVolume and demandShare.

response.monitorSetobject

The prompts a monitor should track, in the order to add them, with coverage of demand, topics and personas.

response.topicsarray

Demand summed by subtopic, with monthlySearchVolume, demandShare and promptCount. personas does the same by audience.

response.brandsarray

Brands people search for in this space, with their monthly searches.

Receive the result

The request returns the task right away with status QUEUED. Poll GET /v1/async/task/{id} until status is COMPLETED or FAILED, or add webhook.url and the finished task arrives at your server with the same task, credits and response. Results stay available for 24 hours after a task finishes.

curl https://api.querying.ai/v1/async/task/$TASK_ID \
  -H "Authorization: Bearer $QUERYING_API_KEY"

Pricing

Each completed research task uses 12 credits.