Building and maintaining automated tests has always been a challenge, especially when it comes to speed, coverage, and collaboration across technical and non-technical roles.
The release of Autonomous 2.2 helps teams reduce the time and effort needed to build and maintain reliable end-to-end tests. It combines LLM-generated test steps, automatic test data generation, built-in API validation, and visual comparisons across environments so that teams can find bugs earlier and release with more confidence.
API Step Builder: Visual API Testing, No Context Switching
Applitools Autonomous has always supported API testing, but with Autonomous 2.2, creating API tests is easier and more accessible than ever.
The new API Step Builder lets you build, run, debug, and update API calls with the UI, directly in your workflow—no code or external tools required.
- Paste in raw cURL or HTTP
- Translate into readable test steps
- Add headers, cookies, auth, body, and parameters visually
- Chain responses into UI steps via variables
Supported Features:
- Full HTTP method support: GET, POST, PUT, PATCH, DELETE, etc.
- Built-in request fields for headers, cookies, authorization, body, and URL params
- Dynamic variables: Ex. use {response.body.token} or {response.body.name} like variables in downstream steps
- Support for raw cURL or HTTP input with automatic translation to English-readable steps
Whether you’re preparing state, validating backend responses, or asserting API behavior, you can now do it all in one place—with a clean, intuitive UI that works just like the rest of Autonomous.
LLM Generated Test Steps: From Business Logic to Automation in Seconds
Write test steps the way you think—in plain English. With LLM Generated Test Steps, you can simply describe your business logic (e.g., “Select Ohio from the state dropdown menu”) and Autonomous will convert it into executable steps.
Let’s say you’re testing a job application page. Instead of clicking through each form field manually, you simply type:
Submit the form as an obscure Tolkien character
Autonomous analyzes the rendered page in real-time, understands the structure and semantics, and breaks down that one line into multiple actionable steps, such as:
- Typing the name and email address
- Selecting an appropriate option from dropdown menus
- Clicking the “Submit” button
The result? A complete, readable, and editable test created in seconds—no coding, no selectors, no extra UI to learn.
Perfect for:
- Speeding up test creation during sprints
- Empowering manual testers to create automated tests
- Eliminating writer’s block in test authoring
🎥 Join our Platform Pulse webinar on July 24 to see Autonomous 2.2 in action
Test Data Generation: Realistic Input with Zero Effort
Test coverage is only as good as the data behind it. Autonomous 2.2 introduces natural language–based test data generation that instantly produces realistic data at runtime, helping with edge case coverage and state variability.
Just say what you need:
Submit the form using random data for 90’s TV dads
Autonomous parses your request, uses LLMs to create diverse, high-quality data, and injects it directly into your test. Each run can generate fresh values—so no two executions are the same.
Cross-Environment Visual Comparisons: One Baseline, Many Environments
Ensuring parity across environments—staging vs. prod, dev vs. QA—or locales is critical. Parameterized visual baselines allow test engineers to define and run tests using variable data while still maintaining deterministic baseline comparison.
For example, you can run the same test in different plans—one with {env}=integration, another with {env}=staging—and compare visual differences across each environment without changing a single line of test logic.
- Parameters and datasets handled automatically
- Custom flows and URL List tests supported
- Visual baselines set per unique parameter set
- Abort prevention and value prompts to ensure test validity
Parameter handling is enforced at both test and plan levels—tests abort cleanly if required values are missing, ensuring reliable execution outcomes.
Tunneling Enhancements: Instant Feedback, Zero Guesswork
Tunneling just got easier and more reliable. With tunneling enhancements, Applitools Autonomous now includes improved diagnostics, real-time tunnel status, actionable error codes, and visibility for team admins—not just account owners
Users now get real-time diagnostics:
- Is the tunnel connected?
- Which hosts is it routing?
- Are there errors, and what do they mean?
Admins now get access to error codes, logs, host mapping, and verification checks—all without reaching out to support.
The Bottom Line
Applitools Autonomous 2.2 turns plain English into production-grade automation. Whether you’re testing UI, APIs, internal systems, or staging vs. production, it removes friction that slows QA teams down.
Explore the new features in the product documentation or dive into Autonomous with a free trial.
Quick Answers
Applitools Autonomous is an AI-powered test automation platform that enables testers to create, run, and maintain end-to-end tests using plain English—no code required. It uses LLMs (large language models) to generate tests and test data automatically, removing the need for coding, locators, or frameworks.
With Cross-Environment Visual Comparison, Applitools Autonomous 2.2 understands what’s important visually. It automatically adjusts for environment-specific differences (like fonts or layouts), so you only get alerted to real UI regressions.
No. Applitools Autonomous understands the test context and structure of your application and test flow. You don’t need to define schemas, write mocks, or manage fixtures manually.
Yes. Autonomous is ideal for automated regression testing, as it can quickly generate test cases that cover expected behavior and catch visual or functional regressions—without manual upkeep.
Yes. You can assert specific response codes, headers, or JSON payload content. For example, “Verify that the response includes a user with ID 1234” will create the necessary assertions automatically.
