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

# 🧪 Chatbot Evaluation

> Test and compare chatbot prompts across multiple AI models

🎥 Refer to the setup video for a step-by-step visual guide:

<div
  style={{
position: "relative",
paddingBottom: "56.25%", // 16:9 aspect ratio
height: 0,
marginTop: "16px",
borderRadius: "14px",
overflow: "hidden",
}}
>
  <iframe
    src="https://www.youtube.com/embed/pGOH__pS3O0?si=gh-ZIxy6V8m30qbV"
    title="Chatbot Evaluation"
    frameBorder="0"
    allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share"
    allowFullScreen
    style={{
  position: "absolute",
  top: 0,
  left: 0,
  width: "100%",
  height: "100%",
  borderRadius: "14px",
}}
  />
</div>

## Overview

The **Chatbot Evaluation** tool allows you to test your prompts across multiple AI models, all at once.

You can compare responses side-by-side, analyze quality, and decide which model works best for your chatbot before deploying changes.

This helps improve accuracy, tone, and user experience without guesswork.

## Accessing Chatbot Evaluation

<Steps>
  <Step title="Open Evaluation Tool">
    * Log in to your dashboard
    * Go to **Settings**
    * Select **Chatbot Settings**
    * Click on **Chatbot Evaluation**
  </Step>

  <Step title="Choose Test Mode">
    * Pick **Single Prompt** or **Multi Prompt** depending on your testing needs
  </Step>
</Steps>

<img src="https://mintcdn.com/neuraltalk/lL8KxG5DfvXuB77p/images/chatbotEvaluation.png?fit=max&auto=format&n=lL8KxG5DfvXuB77p&q=85&s=f5e33e24335791dbb9c9277cf0692d1a" alt="Chatbot Evaluation Page" width="1482" height="769" data-path="images/chatbotEvaluation.png" />

## Evaluation Modes

### Single Prompt

Test one prompt at a time and compare responses instantly.

### Multi Prompt

Run multiple prompts in sequence to analyze consistency across AI models.

> Use **Multi Prompt** when testing workflows, FAQs, or repetitive scenarios.

## Selected Chatbot

You can evaluate prompts using your existing chatbot configuration.

<CardGroup cols={2}>
  <Card title="Chatbot" icon="robot">
    Example: **ABC**
  </Card>

  <Card title="Active Model" icon="cpu">
    Example default: **gpt-4o-mini**
  </Card>
</CardGroup>

The chatbot’s greeting and behavior settings are applied automatically during testing.

## Comparing AI Models

The evaluation area displays each selected AI model with its response.

Each response panel includes a **Create Correction** option.

### Create Correction

Use this to:

* Suggest improvements
* Fix tone or clarity
* Adjust formatting or accuracy

Corrections help refine how the chatbot learns and responds over time.

## Selecting AI Models

### Select AI Models

Choose which models to include in the comparison.

<CardGroup cols={2}>
  <Card title="Add or Remove Models" icon="list-plus">
    Click **+ Add Model** to include additional AI engines.
  </Card>

  <Card title="Flexible Testing" icon="shuffle">
    Mix different providers or versions for deeper evaluation.
  </Card>
</CardGroup>

## Saving Changes

When you're satisfied with your testing results:

<CardGroup cols={1}>
  <Card title="Save Changes" icon="save">
    Apply updated model selections or evaluation settings.
  </Card>
</CardGroup>

## Best Practices

* Test prompts for both short and long responses
* Compare tone, accuracy, and consistency across models
* Use corrections regularly to fine-tune results
* Re-evaluate after major chatbot updates
* Document findings for your team