camelQA
CamelQA, Inc. · Coding
camelQA was a Y Combinator-backed AI QA agent that turned plain-English test cases into automated runs on real iOS and Android devices, with no source code needed. The company no longer sells it.
About camelQA
What Is camelQA
camelQA was an AI agent built for mobile app testing. You described a test in ordinary words, the way you would explain it to a colleague, and the agent carried it out on a real phone: opening screens, tapping buttons, typing into fields, and checking what happened next.
The pitch aimed at a familiar pain. Writing the first hundred automated tests is manageable. Keeping them alive while the interface changes every sprint is what breaks QA teams, and a single renamed button can take an entire suite offline.
camelQA attacked that problem with computer vision instead of element selectors, one of the earlier attempts at automated mobile testing that ignored the element tree entirely. The agent read the screen the way a person does, which meant it kept working when the layout shifted and it never needed your source code.
The product is no longer sold. camelqa.com now redirects to camelai.com, where the same team ships camelRun, camelBot and camelCode, and no sign-up or changelog for the QA agent remains. Treat this page as a record of what the tool did, not a live buying guide.
Getting Started
- Create an account and hand over a build of the app you want tested. No SDK, test framework or source code access was involved.
- Write the test case in plain English, describing the goal rather than a list of taps and swipes.
- Choose the platform and device for the run, then start it.
- Read the result, then keep the case and run it again on the next build.
Product Information
A quick look at camelQA's pricing, supported platforms, and performance.
Best for
The users, tasks, and scenarios where this tool fits best.
Users
- QA leads at small mobile teams
- Product managers without automation experience
- Founders shipping weekly builds
Tasks
- Smoke testing a release candidate
- Regression checks on core journeys
- Accessibility checks
Scenarios
- A redesign breaks half the suite
- Nobody wants to own Appium upkeep
- The night before an app store submission
Key features
Natural language test cases
Test cases were written as goals, not scripts. You typed what a user should be able to do and the agent worked out the steps, which meant a designer or a support lead could contribute cases without touching code.
Vision-based screen reading
camelQA combined accessibility element data with an in-house object detection model and Google SigLIP for classifying UI elements. In practice the agent recognized a button by how it looked and what it was labeled, so a moved control did not automatically fail the test.
Runs on real iOS and Android devices
Tests executed on physical phones rather than emulators, which matters for anything touching camera, notifications or touch latency. iOS and Android testing shared one workflow, and neither needed source code or special build flags.
Self-correcting during a run
When a screen did not match expectations, the agent adjusted its next step instead of stopping dead. That behavior is the difference between a test that reports a real bug and one that reports a renamed label.
More than end-to-end flows
The team scoped the agent for end-to-end, smoke, sanity and accessibility testing. A single case list could cover a full purchase journey and a two minute check that the app still opens, which kept the suite small enough to run often.
Pros and cons
Pros
- Test cases read like instructions, so the whole team could write and review them.
- Vision-based element detection cut the maintenance work that makes mobile test suites rot.
- Real device runs caught issues that emulators hide, including timing and permission problems.
- No source code or SDK integration was needed, which shortened setup to minutes.
Cons
- Mobile only. There was no web or desktop coverage, so a cross-platform product still needed a second tool.
- The product is retired. There is no sign-up page, no public pricing and no support channel today, which rules it out for new projects.
- Pricing beyond a free allowance of 10 test cases was never published, so budgeting was a conversation rather than a lookup.
Frequently asked questions
It ran mobile app tests from plain English descriptions. You named a goal, the agent executed it on a real iOS or Android device, and you got a pass or fail result without writing automation code.
Related content
Explore related tools, skills, and articles for camelQA.
camelQA Alternatives
Base Code
Base44 (Wix) · Coding · ProductivityBase Code is a shared cloud development environment from Base44. Teammates describe changes to an existing GitHub codebase in chat, preview them live and send them back as pull requests.
Google EmbeddingGemma 2
Google DeepMind · Writing · Image · CodingGoogle EmbeddingGemma 2 is a 740M-parameter open model that turns text, code, images, audio, and video frames into vectors in one 768-dimension space, running on phones and laptops.
GitGlow
GitGlow · Coding · ProductivityGitGlow is a local-first desktop Git client that lets independent developers review their own code and their AI agents' changes in one place before shipping.