What is the Test Deduplication Agent?

The Test Deduplication Agent is the AI duplicate scan built into TestMu AI Test Manager. It reads the manual test cases in a project or folder, groups the ones that describe the same check into duplicate sets, scores each match from 0 to 100, and lets you compare and delete the copies you do not want. Nothing is deleted until you mark the cases and confirm.

How does the agent detect duplicate test cases?

It uses semantic matching rather than exact text matching. Each test case is converted into a vector embedding built from its title, steps, and description, and the scan runs a nearest-neighbour similarity search across the scope you picked. Two cases that use different words for the same steps still land in the same group, with a percent match that reflects how close they are.

What is a base test case?

The base test case is the one the rest of a group is compared against, and the one that survives when you delete its duplicates. A duplicate is any case that scored above your minimum match score against that base. If a duplicate is better written or more current, open the menu on the base case and promote the duplicate; the group re-scans against the new base at no extra credit cost.

What does a deduplication scan cost?

A scan costs a flat 5 AI credits at any project size. Credits are charged only when the scan produces at least one duplicate group; a scan that finds nothing costs nothing. The scan settings, credit rules, and limits are documented in the test case deduplication guide.

Can I undo a deletion?

No. Deleting duplicates is permanent and there is no restore. That is why the workflow separates marking from committing: marked cases persist while you move between groups, the footer shows the total by match band, and the final dialog requires you to type DELETE before anything is removed. Test runs that included a deleted case stay in place.

Does it work on automated test scripts?

No. The agent scans manual test cases only, including cases authored with KaneAI and standard manual cases. Automation scripts in the project are never read or touched by a scan.

Can it compare test cases across projects?

No. A scan runs inside one project, either the whole project or a single folder with its subfolders, and only one scan can run per project at a time. For large repositories, scanning folder by folder keeps each result set reviewable.

Why are some match scores unreliable?

The embedding is built from the title, steps, and description together. A test case that has only a title gives the model very little to compare, so its score can be misleading in either direction. Fill in steps and a description before scanning, or treat title-only cases as candidates to review rather than results to trust.

How is this different from the Smart Context Memory Layer?

They work at different moments. The Smart Context Memory Layer runs during AI test case generation and steers the generator toward reusing an existing case instead of writing a near copy. The Test Deduplication Agent runs on demand against cases that already exist, which is what you need after an import, a migration, or months of manual authoring.

Which plan includes the Test Deduplication Agent?

Scans draw from the AI credit balance on your account. Test Manager Premium is $49 per user per month billed annually and includes 2,000 AI credits each month, enough for hundreds of scans alongside AI test case generation. Current plans are on the pricing page.

Find and Remove Duplicate Test Cases With AI

The Test Deduplication Agent scans your Test Manager project, scores every semantic match from 0 to 100, and lets you compare before you delete.

Start free with Google

Start free with Email

Automate Browser Flows from your
Terminal with Kane CLI

Explore Kane CLI

Trusted by 3M+ users globally at

+Read case study

The Test Deduplication Agent Shows Its Work

Pick a scope, get every duplicate scored against a base test case, filter the list your way, and compare before anything is deleted.

Start free with Google

PROJECT OR FOLDER SCOPE

Scan a Whole Project or One Folder

Choose Whole Project to review every folder, or Specific Folder to clean up one area with its sub-folders. Set the minimum match score from 1 to 100.

Scan a project free

Whole Project covers all folders and sub-folders
Specific Folder scans one folder plus its sub-folders
Title-only cases scan too, with less reliable scores

MATCH SCORES

Every Duplicate Is Scored Against a Base

Each group has one base test case that stays. Every duplicate shows a percent match to it, so you keep the base and delete only the copies.

See it in Test Manager

Base test case is kept; marked duplicates are deleted
Percent match on every duplicate, sorted high to low
Change the base when a duplicate is the better version

FILTERS

Filter Duplicates the Way You Work

Narrow a group by match band, then by PR status, status, tags, automation status, priority, type, issues, or who created the case and when.

Filter duplicates free

Match bands: above 80, 50 to 80, 20 to 50, below 20
PR status, status, tags, and automation status filters
Created at, created by, issues, priority, and type

COMPARE AND DELETE

Compare, Mark, Then Delete

Compare opens both cases with an AI summary and a Diff, Base, or Duplicate view across Steps, Details, and Linked Items. Mark what goes, then confirm.

Compare duplicates free

AI summary says what differs, or that nothing does
Ignore drops a case from the group without deleting it
Typed DELETE confirmation; test runs stay in place

5 credits

per scan, at any project size

0 credits

when no duplicates are found

1 to 100

adjustable minimum match score

4 bands

of match confidence to filter by

Three Ways to Keep a Test Suite Free of Duplicates

Use the agent to clean up existing cases, the memory layer to stop new copies at generation, and reviewer judgment for the final call.

Test Deduplication Agent

Smart Context Memory Layer

QA Lead Review

When it runs

On demand, after an import or a growth spurt

During AI test case generation

During suite audits and release planning

What it compares

Title, steps, and description as embeddings

New prompts against existing cases

The cases the reviewer opens side by side

Output

Scored groups with a diff and an AI summary

Reuse of an existing case instead of a new one

Keep or delete calls with reviewer context

Scope

One project or folder, manual cases

The project being generated into

The folders the reviewer owns

Cost

5 credits per scan that finds results

Included in generation

Reviewer time

When a Dedup Scan Pays for Itself

Start free with Google

Duplicates pile up at predictable moments. Run the agent right after them.

After a 1-click migration

Imports from a legacy test management tool carry years of copied cases. Scan the imported folder first.

After sprints of manual authoring

Testers writing cases by hand across sprints and folders end up covering the same flow twice. One scan surfaces the overlap.

Before a release cycle

A lean suite means fewer cycles to assign and truer coverage numbers on the dashboard.

Part of the TestMu AI Quality Engineering Platform

Test Manager sits beside AI authoring, execution, real devices, and insights, so a clean suite feeds every stage of the pipeline.

Test Manager

KaneAI

HyperExecute

Real Device Cloud

Agent Testing

Test Intelligence

Test Manager

Plan, organize, and report on manual and automated testing in one workspace, with AI writing the test cases.

More about Test Manager

AI Test Case Generation

Structured cases from natural language, Jira tickets, PDFs, and images.

Two-way Jira and ADO Sync

Defects flow out with full context; status syncs back to the case.

1-click Migration

Move a whole repository across in one click without rebuilding it.

Success Stories of TestMu AI (Formerly LambdaTest)

50%

reduction in test execution time

“HyperExecute is a highly reliable test execution platform and has excellent customer support.”

Sagar Uday Kumar

Sr. Engineering Manager

Some Love from our Customers

As Best Egg expanded its product offerings and entered new markets, we knew our old testing infrastructure couldn’t keep up.
With support from Tenny Agustin, our Engineering Operations Lead, we modernized our approach with

TestMu AI

Best Egg

best-egg

Excited to Share My Learning Journey with Kane AI & Lambda Tool!
I'm pleased to announce that I've recently gained hands-on experience exploring Kane AI through the Lambda Tool and it’s been a fantastic journey of upskilling!

KaneAI

Suryateja Goud

suryateja-goud

See how is #Futureready to enable blazing-fast test orchestration seamlessly integrated with organizations' existing CI/CD platforms, using #Microsoft Azure.

TestMu AI