QF-Test 11.1 – Improved AI test generator and more

Hi! My name is Daniel, and today I’ll show you what’s new in QF-Test 11.1.

With QF-Test 11, we brought AI into your test automation. Version 11.1 builds right on that: the AI now generates real, maintainable QF-Test tests, and QF-Test supports your agents even during development. I’ll also show you new features in the UI Inspector.

AI assistant: test generation

Let’s start with the test case generation assistant you know from QF-Test 11. Generation is now much faster and runs continuously: the assistant creates test after test until you stop it.

Wherever possible, it creates “real” QF-Test nodes directly instead of AI Instructions, including checks. The result: tests that run fast and reliably, without any AI at all.

Extending an existing suite

And you don’t have to start from scratch. The assistant analyzes and extends your existing test suite and only inserts new nodes instead of creating a whole new suite. If you interrupted the generation, simply pick up where you left off later.

Your own procedure library

Especially valuable: the assistant works with your own procedure library via procedure call recording. The generated tests then call your existing procedures with procedure call definitions instead of recording everything from scratch. That way, generated tests fit your project and stay easy to maintain.

Converting AI Instructions

And your existing AI Instructions? You can now replay them with one click and record them at the same time. Wherever possible, QF-Test automatically converts the AI interactions into reliable, classic nodes.

From the outside, too: via MCP

All of this also works from the outside. Since QF-Test 11, you can control QF-Test from your own terminal or IDE via the integrated MCP server.

Now there are new tools for editing suites: your agent, for example Claude Code, records its interactions with the application directly as tests or procedures in a test suite.

And, true to our “bring your own model” approach, you can now add your own CLI-based AI providers alongside Claude Code and GitHub Copilot.

Dev-Connect: QF-Test during development

The next big step: QF-Test during development. Your AI agent starts the application straight from your development environment, using your build script or run configuration. QF-Test connects automatically, without the application having to be started from QF-Test.

This lets the agent check every change immediately in the running application. This works with Java and web applications, too. That makes QF-Test the interface for fully agentic development: write code and check it, in a loop.

UI Inspector: search and filter

But enough about AI: robust UI test automation remains at the core of QF-Test. One tool you use for it every day is the UI Inspector. It now offers search and filtering for the component tree, for all supported technologies.

• The search highlights matching components.
• With filter chips, you reduce the tree to exactly those components and their parent elements, so you keep the context.
• You set the search scope separately for search and filter: tree text, technical details or QF-Test-specific details.
• Matches are highlighted in the detail view as well.

So even in large UIs, you find what you’re looking for in seconds.

More new features

Also new in 11.1:

• QF-Test starts noticeably faster
• Checks can store the actual value in a variable, e.g. for custom error handling

And there are many more features and improvements in Version 11.1.

---

In summary

With QF-Test 11.1, the AI generates real, maintainable tests, extends your suites and uses your own procedures. Your agents can now use QF-Test during development, and the UI Inspector finds anything in seconds.

---

As always, you’ll find all the new features in the detailed release notes on qftest.com. There’s also a free trial, so you can get started with QF-Test 11.1 right away.