Understanding how people search with AI
Search behavior is changing. People are no longer relying only on traditional search engines like Google and Bing to find information. Many now turn to AI tools like ChatGPT and Gemini to ask questions, compare solutions, and explore topics in a more conversational way.
The AI Models Report in AnswerThePublic helps you understand those interactions by showing prompts people use in AI conversations.
Instead of focusing only on keywords, the report helps you understand how people phrase questions in AI conversations and what kind of answers they expect.
In practice, this helps you identify content opportunities, understand emerging search behavior, and create content that aligns more closely with how people interact with AI today.
How the AI Models Report works
The AI Models Report focuses specifically on prompts for AI platforms such as ChatGPT and Gemini.
To start exploring the data, open the AI Models section from the left-hand menu, enter a keyword, choose your language and target location, and click Search AIs.
The report then gathers prompts related to your topic and organizes them into a visual report that helps you spot patterns and opportunities more quickly.
Unlike traditional keyword reports, the focus here is not just search volume, but the way people communicate with AI tools and the intent behind those conversations.
Understanding the AI Models report
The keyword wheel is the most visual way to explore AI prompts related to your topic.
Prompts are grouped by search intent — Informational, Navigational, Commercial, and Transactional — helping you quickly understand the types of conversations happening around your keyword.
This makes it easier to identify what users are actually trying to accomplish when interacting with AI tools.
You can click prompts directly from the wheel to open the Table View, where you’ll see all related prompts in a structured list format.
Clicking on an individual prompt opens the Prompt Overview, which provides additional context around the conversation.
Depending on the prompt, you may see:
- The AI source (ChatGPT or Gemini)
- Search intent
- Average sentiment
- Brands mentioned
- AI-generated responses
- Response dates
The report may also highlight brand mentions inside prompts, helping you understand how users compare products, ask for recommendations, or reference competitors in AI conversations.
For many prompts, you’ll also see the option to Get AI Response. Some insights, such as sentiment visibility and brand references, become more detailed after a response is generated. This allows you to generate a response based on the selected prompt, so you can better understand how AI tools may answer that question.
This can help you identify content gaps, understand how competitors are being mentioned, compare AI-generated answers with your own content, and explore how topics are framed inside AI-driven discovery.
The number of AI Responses you can generate each month depends on your subscription plan. For example, Starter plans include up to 30 AI prompt responses per month.
This means Starter users can click Get AI Response up to 30 times during each monthly cycle.
These limits reset automatically on the first day of the month. If you reach your limit, you’ll see an option to upgrade your subscription for additional access.
Prompt visibility also depends on your plan. For instance, Starter users can view up to 1,000 prompts per month.
You can find more details in the dedicated Limits article.
Understanding search intent and sentiment
The AI Models Report includes more than just prompts. It also provides context that helps you understand why people are asking certain questions and how they feel about a topic.’
Search intent explains the goal behind a prompt. Some users are trying to learn something, others are comparing solutions, looking for a specific website, or preparing to make a decision. For example, a prompt like “what is technical SEO” would typically be informational, while “best SEO tools for agencies” would lean commercial.
The report organizes prompts into four main intent categories:
- Informational
- Navigational
- Commercial
- Transactional
Understanding search intent is important because it helps you create content that better matches what users actually expect to find.
For example, someone searching with informational intent usually expects education or explanation, while commercial and transactional prompts often reflect users who are evaluating options or getting closer to taking action.
This makes it easier to create content that feels more relevant and useful, both for traditional search engines and AI-driven discovery.
The report also includes sentiment, which helps identify whether prompts are generally positive, neutral, or negative.
Negative sentiment may reveal frustrations, objections, or problems users are trying to solve. Positive sentiment can indicate interest, curiosity, or strong purchase intent, while neutral sentiment is often associated with exploratory or educational searches.
While sentiment is not a direct ranking factor, it provides useful context around tone and user expectations. In practice, this can help you create clearer, more aligned content that better reflects the way people naturally ask questions.
How to use AI prompt insights strategically
The AI Models Report works best as a research and discovery tool, not just a keyword report.
One of the most effective ways to use it is to identify how people naturally phrase questions in AI conversations. These prompts are often longer, more specific, and more contextual than traditional searches, making them useful for creating educational content, FAQs, tutorials, and in-depth articles.
The report is also valuable for spotting emerging topics earlier. Because AI conversations often surface exploratory questions quickly, they can reveal opportunities before they become highly competitive in traditional search results.
You can also use the report to understand how brands and products are discussed inside AI-generated conversations. This can help you identify positioning opportunities, competitor mentions, and gaps where your content could provide more complete or helpful answers.
In practice, the goal is not simply to copy prompts into content, but to better understand how audiences interact with AI tools and what information they expect to find.
As you explore opportunities, you can also refine the report using filters like Search Intent and Sentiment. This makes it easier to focus on specific types of conversations, whether you’re researching educational content, commercial intent, or prompts that reveal frustrations, objections, or strong purchase interest. Instead of reviewing every prompt manually, filters help you narrow the report down to the opportunities that matter most for your strategy.
How this fits into your content strategy
The AI Models Report works best when combined with the rest of AnswerThePublic.
A common workflow is to identify opportunities inside AI Models, validate broader search demand using the other search tabs (All, Search Engines, Social Media, Shopping) and People Also Ask data, and then move into Content Studio to generate content around those themes.
Instead of looking only at keywords and rankings, the report helps you understand how people ask questions, compare solutions, and interact with AI tools like ChatGPT and Gemini in real conversations.
As AI search behavior continues to evolve, these insights can help you adapt your content strategy earlier and stay aligned with how people discover information online.