This content originally appeared on DEV Community and was authored by Ranit Mukherjee
As a developer, I'm no stranger to API testing. My typical workflow involves jumping between my code editor, Swagger docs, and a terminal window where I carefully craft curl commands or set up requests in Postman. It’s a necessary part of the job, but it’s also repetitive and time-consuming.
So, when I heard about Keploy, an AI-powered tool that promised to automate this whole process, I was skeptical but intrigued. Could it really turn my manual testing grind into a few simple clicks? I decided to find out by testing it on two fun, public APIs.
The Challenge: Testing Two APIs in a Few Minutes
To give Keploy a fair shot, I picked two different kinds of APIs:
- Wizard World API: A fun and complex API with lots of different endpoints for getting information about wizards, spells, elixirs, and magical houses.
- Emojihub API: A very simple API that does one thing: gives you a random emoji when you call it.
My goal was to see how quickly I could generate meaningful tests for both using the Keploy Chrome Extension.
Test 1: Exploring the Wizarding World
First, I tackled the Wizard World API. Instead of writing curl requests for each endpoint, I just followed these steps:
- I installed the Keploy Chrome Extension.
- I navigated to the Wizard World API's documentation page, which has interactive endpoints.
- I clicked "Start Recording" in the Keploy extension.
- I browsed the API like a normal user, clicking to get a list of
Wizardsand then fetching a fewSpells. - After about 30 seconds, I hit "Stop Recording."
Instantly, Keploy showed me what it had captured. It had recorded each API call I triggered and automatically converted them into complete, ready-to-run test cases. The requests, headers, and full JSON responses were all there. No manual setup, no typing, just a few clicks.
Test 2: Getting a Random Emoji
Next up was the much simpler Emojihub API. The process was just as easy:
- I went to the Emojihub API endpoint for a random emoji.
- With Keploy recording, I simply refreshed the page a few times.
- Each time I refreshed, the browser called the API to get a new emoji, like
{"name":"old woman, type-4",...}. - Keploy captured every single one of these calls.
In seconds, I had a series of test cases for an endpoint that produces random results. Manually, this would have been a minor but still present chore. With Keploy, it was effortless.
The Verdict: Why I'm Sold on AI-Powered Testing
My short experiment was a huge success. The difference between my old manual process and using Keploy was night and day.
| Task | The Old Manual Way | The New Keploy Way |
|---|---|---|
| Test Creation | Manually write curl commands or configure Postman for each endpoint. |
Click around the app while Keploy records in the background. |
| Effort | High. Requires typing, copying, and pasting. Prone to typos. | Minimal. Just browse the application as a user would. |
| Time Spent | Minutes per endpoint, adding up to hours for a full API. | A few seconds per user flow. |
| Test Realism | Tests the API in isolation. | Tests are based on actual user interactions. |
Moving from manual methods to an AI-driven tool felt like a superpower. I went from the tedious task of writing tests to simply using my application and letting the tests write themselves. For any developer looking to increase their test coverage without slowing down, a tool like this isn't just a nice-to-have; it's a game-changer. It frees you up to do what you do best: build great software.
[1] https://wizard-world-api.herokuapp.com/swagger/index.html
[2] https://emojihub.yurace.pro/api/random
[3] https://swagger.io/docs/
This content originally appeared on DEV Community and was authored by Ranit Mukherjee
Ranit Mukherjee | Sciencx (2025-06-28T00:54:56+00:00) API Testing: From cURL to Click. Retrieved from https://www.scien.cx/2025/06/28/api-testing-from-curl-to-click/
Please log in to upload a file.
There are no updates yet.
Click the Upload button above to add an update.