This page explains how to use various generative AI features in MagicPod, along with a list of the applicable features.
Table of Contents
- How to Enable Generative AI Features
- AI credits
- Currently Available Features
- Frequently Asked Questions
How to Enable Generative AI Features
To enable generative AI features, go to your organization settings page, open the "Optional Features" tab, and select “Enable” under Generative AI Features, then click "Update".
This setting is available to any user with update permissions for the organization.
Additionally, we offer a setting for Enterprise plan users that allows the generative AI features to be enabled on a per-project basis. For more details, please refer to the help page: Controlling generative AI features per project
AI credits
AI credit is a credit system required to use certain generative AI features (as of December 2025: MagicPod Autopilot for all users and AI Assertion for Old Standard Plan subscribers). AI credit has a free usage tier (monthly and initial). Once you've used up your free tier, you can purchase paid credits to continue using certain generative AI features.
Please refer to the help page: AI credits
Currently Available Features
The list of features covered by the Generative AI toggle is updated as needed.
Please take a look at this page for the latest information. Note that you cannot enable or disable features individually. Turning on Generative AI Features will apply to all supported functionality.
Users' inputs are never used to train the generative AI models in any feature.
Test Case Summary
When you create a test using MagicPod, a summary of the test’s purpose and steps is automatically generated.
Previously, when a different team member needed to understand someone else’s test case, they had to read through the steps or ask the creator directly. This could be time-consuming and makes it difficult to grasp the overall intent.
With automatic summaries, even team members unfamiliar with the test can quickly understand its purpose by reading the summary, making knowledge sharing much easier.
Additionally, summaries are automatically updated when a test is modified, so there’s no need to update them manually or worry about them becoming outdated.
Test Change Summary
When saving changes to a test case, you can generate a suggested change comment with one click.
The suggested comment can be edited as needed before saving.
Previously, writing clear change messages was a manual task that took time and effort, especially to maintain an easy-to-follow change history. By using generative AI, this process becomes much faster. It also encourages more users to write meaningful change messages, improving the clarity and usefulness of the test history.
Use of MagicPod Web API via the MCP Server
With the MagicPod MCP server, you can issue natural language commands to an AI agent and automate MagicPod operations.
The MCP server supports calling MagicPod Web APIs and searching help pages. Among these, enabling Generative AI Features is required to call the Web API.
MagicPod Autopilot
MagicPod Autopilot is a test automation agent that leverages AI technology. It enables test case creation and editing through natural language input. It simplifies the traditional test creation process, allowing both technical and non-technical users to create test cases intuitively. It also includes a failure analysis feature where MagicPod Autopilot automatically investigates and analyzes the cause when a test fails.
⚠️Using MagicPod Autopilot requires AI credit.
AI Assertion
AI Assertion is the feature that allows generative AI to understand screen and text content and automatically perform necessary checks simply by describing "what results you expect" in natural language within your test case.
⚠️For old Standard Plan subscribers, using AI Assertion requires AI credit.
Use of Generative AI by the MagicPod Support Team
To help resolve customer inquiries more quickly, the MagicPod support team may pass inquiry texts and attachments sent from users to generative AI.
Frequently Asked Questions
Are input contents and test case information used for training generative AI?
User input contents and test case information are not used for training generative AI models. Additionally, MagicPod does not use customer test case data for training purposes either.
Test cases are not being summarized even after enabling the Generative AI features
AI-generated summaries are created every hour. Please note that summaries will not be generated immediately after enabling the feature or right after creating/updating test cases.
What generative AI models are being used?
As of November 2025, the following models are being used. However, these may be changed without prior notice. When making changes, models that do not use user input for training the generative AI models will be used.
- Test Case Summary feature: GPT-4o mini
- Test Change Summary feature: GPT-4o mini
- MagicPod Autopilot: Claude Sonnet 4.6
- AI Assertion: Claude Haiku 4.5
- Failure analysis feature: Claude Haiku 4.5