> ## Documentation Index
> Fetch the complete documentation index at: https://docs.geoark.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Prompts

> Track the questions people ask AI — and see when your brand shows up in the answers.

## Video walkthrough

<iframe className="w-full aspect-video rounded-xl" src="https://www.youtube.com/embed/k2QdInXzslc" title="GeoArk AI - Prompts" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowFullScreen />

## What Are Prompts?

**Prompts** are the questions you want to track — the same questions people type into ChatGPT, Claude, Gemini, or Perplexity. Examples:

* "Best project management tools for startups"
* "Recommend a CRM for small business"
* "Top restaurants in Austin"

When you **run** a prompt, GeoArk sends it to the AI platforms you choose and shows whether your brand appears in the answer — and if so, where you rank.

## How to Create a Prompt

1. Go to **Prompts** in the sidebar
2. Click **Add prompt** or **Run prompt**
3. **Select a brand** — The brand you want to track
4. **Enter the query** — Type the question (e.g., "Best CRM for small teams")
5. **Choose AI models** — Pick which platforms to run on (your plan controls which are available)
   * **Free plan:** Gemini only
   * **Starter+:** All 4 (ChatGPT, Claude, Gemini, Perplexity)
6. **Schedule** — Choose:
   * **Once** — Run now only (Free plan only option)
   * **Daily** — Run every day at 9 AM (Starter+)
   * **Weekly** — Run every Monday at 9 AM (Starter+)
7. Click **Run prompt** (or **Run now & schedule** if you chose Daily/Weekly)

## How to View Results

After a prompt runs:

1. The prompt appears in the list with status **Completed**
2. Click the prompt to open the details
3. You’ll see:
   * **Brand mentioned** — Yes/No per platform
   * **Visibility score** — 0–100
   * **Position** — Where your brand ranked (1 = first)
   * **Entity type** — Whether the AI surfaced you as a **Person** or **Organization** (platform bias: Perplexity often favors individuals, ChatGPT favors firms)
   * **Review language** — Our AI agent extracts semantic phrases models used (e.g., "highly regarded", "leading provider")—how you're described, not just if you're cited
   * **Full AI response** — The actual answer from each model

You can switch between platforms (ChatGPT, Claude, etc.) to compare responses.

## Re-running a Prompt

Click the **Run** (play) button on a prompt to run it again. Useful for checking if visibility has changed.

## Prompt Limits

* **Free:** 2 prompts per month, manual run only, results visible 14 days
* **Starter:** 50 prompts per month
* **Growth:** 200 prompts per month
* **Custom:** Custom limits

One prompt counts as one use, even if it runs on multiple platforms.