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

# AI tests and results

# Run AI tests and read prompt results

AI tests send a saved prompt to supported AI models and store each model response for analysis.

<img src="https://mintcdn.com/tamlr/nVgF79AyVUKqqUbo/kb/screenshots/app-prompt-detail.png?fit=max&auto=format&n=nVgF79AyVUKqqUbo&q=85&s=cd18ab9fc4b2bad880bfb6439e110dce" alt="Prompt detail" width="1087" height="729" data-path="kb/screenshots/app-prompt-detail.png" />

## Use cases

* See which AI models mention the brand.
* Compare answers from ChatGPT, Gemini, Perplexity, Claude, and Google AI Mode where configured.
* Review sentiment, citations, competitor mentions, and search ranking context.
* Rerun a prompt after changing the prompt or company settings.
* Generate optimization actions from weak or missing visibility.

## Open a prompt result

1. Go to **AI Search**.
2. Select a prompt row.
3. Review the prompt detail page.

## Run AI tests

1. Open a prompt detail page.
2. Select **Run AI Tests**.
3. Wait for model responses to complete.
4. Review the updated visibility and model result sections.

## What each result means

* **AI Tests**: number of model responses stored for this prompt.
* **Avg Visibility**: how often the company appeared in the responses.
* **Status**: whether the prompt is active.
* **Model cards**: per-model visibility and response counts.
* **Google search rankings**: keyword ranking results associated with the prompt.
* **Sentiment breakdown**: how positive, neutral, or negative the stored answers are.
* **Brand visibility**: whether the company was mentioned relative to other brands.
* **Sources**: citations extracted from AI responses.

## View a response

Open an individual AI response when you need to inspect the full answer, source citations, brand mentions, or sentiment details.

## Optimization plans

When available, optimization actions help translate weak visibility into content or source-improvement work. Use these actions as guidance for what to publish, update, or promote.

## What Tamlr saves from a test

* The prompt that was tested.
* The AI model that answered.
* The full answer text.
* Whether your company and tracked competitors were mentioned.
* Source citations found in the answer.
* Sentiment and visibility signals used in dashboards and reports.

This saved history lets your team compare results over time instead of relying on one-off AI searches.