You build software for customers running all kinds of different versions, configurations, and deployment setups. Before every update, someone tests it by hand — on your side, and again on the customer's. That ties up staff, costs real money, and burns time you and your customers would rather spend on your actual business. GUI Test Studio takes that off your plate: record it once, and it reliably re-checks every update in minutes — work that takes hours by hand today.
From recording to the finished report, GUI Test Studio covers the entire GUI-testing workflow — with no scripting required.
A global mouse/keyboard capture automatically detects clicks, double-clicks, drags, scrolling, key presses and typed text, turning them into named test steps — including an image anchor around the click point plus the window and control name, so it stays clear what happened even without the recording open.
Recorded test cases run again automatically — with configurable retries on failure, tolerant image comparison, and "wait for image state" for elements that only appear after a delay. Any test case can also resume from an arbitrary step with one click.
Instead of only comparing pixels, Claude judges a screenshot against a freely worded instruction — e.g. "The dialog shows no error message and all required fields are filled in" — and returns a verdict with reasoning. On request, the AI also explains why a failed image checkpoint failed.
During every test run, CPU usage, RAM, thread/handle/module counts, the number of sub-processes, and the target process's network traffic (via ETW) are continuously captured — live in the status bar and an optional overlay that follows the cursor, with configurable thresholds per metric.
If the monitored target process restarts itself during an ongoing monitoring session (same name, new PID), GUI Test Studio automatically reattaches — the live session and metric history stay intact. A plain window switch under the same PID is detected separately, without resetting the monitoring.
.NET exceptions are captured with a real, symbolicated stack trace (via ClrMD), the target application's log/crash directory is continuously scanned for error patterns, and window/navigation and focus changes are logged automatically.
Every run automatically produces an HTML report with screenshots, diff images for failed image comparisons, and any AI reasoning — one click from the app opens it in the browser.
Live sessions can additionally be captured as video — the screen and, optionally, the speaker via webcam including microphone audio as a second video. Both play back in sync right inside the session history, or export as a standalone HTML file with a draggable, hideable picture-in-picture overlay.
An ongoing monitoring run — metrics, diagnostic events and screenshots — can be saved as a named live session at any time and later browsed on a timeline. A ZIP export bundling all referenced images and videos makes it easy to hand off to colleagues or a customer.
No script, no coding required. GUI Test Studio records what you do and repeats it reliably — even after layout shifts in the next update.
Pick the target application and set the starting position for recording & playback (monitor, fullscreen or a fixed resolution).
Use the application as usual — clicks, input and image checkpoints automatically become named test steps.
Run the test case automatically — with retries on failure and robust targeting via image anchors.
The live log, metrics, diagnostic entries and HTML report show exactly what happened — and where it didn't.
Some applications restart themselves after a crash or an update — with a new process ID but the same name. GUI Test Studio detects this automatically and seamlessly reattaches the ongoing monitoring to the new process instead of aborting the session.
Test cases can be organized in arbitrarily nested folders and moved via a target-folder dialog — handy once a project grows to dozens of test cases over time.
Instead of starting a test case by hand over and over, it can repeat automatically 1 to 999 times in a row — with a live progress indicator, ideal for stability and load testing across many runs.
GUI Test Studio is built for robust, repeatable test runs — with warning thresholds, diagnostic depth and project management that go beyond plain replay.
On top of live diagnostics, the screen can be recorded as video at the press of a button — independent of automatic session capture, via its own floating control bar with start/stop. Optionally, the webcam simultaneously records the speaker including microphone audio as a second video.
During recording, a small preview window shows the speaker their own camera image for orientation — technically excluded from the screen capture, so it never shows up in the recorded video itself.
Screen and speaker video play back in sync in the embedded player of the session history — or export as a standalone HTML file where the speaker overlay can be dragged freely and hidden as needed.
The timeline of a saved live session makes every captured event directly clickable, complete with timestamp, metric snapshot and screenshot — ideal for retracing a reported incident in detail after the fact.
Each project can hold separate RAM thresholds for 32-bit and 64-bit target processes. The bitness of the monitored process is detected automatically — once current RAM usage exceeds the matching threshold, the live display turns red instantly.
A freely worded instruction is enough — Claude judges the screenshot and returns a verdict with reasoning, right there as a test step in the log.
ClrMD captures .NET exceptions from the target application with a symbolicated stack trace — complemented by log/crash directory monitoring with custom search terms, and automatically detected window, navigation and focus changes. An optional, experimental debugger attach additionally catches native exceptions.
Click targets are re-found via an image patch around the recorded point rather than fixed coordinates alone — patch size and minimum match score can be fine-tuned in settings.
GUI Test Studio ships as a standalone x86 and x64 build — important for accurate ClrMD diagnostics, since a bitness mismatch between the test tool and the target process prevents reading memory information at all.
While a test case runs, the step currently executing highlights in pastel blue in the test-step list; completed steps stay marked green or red depending on the outcome — all semi-transparent, so thumbnails and description text stay readable underneath.
A GridSplitter between the main area and the test-run log can be dragged to resize — handy when you need to keep an eye on many log or diagnostic lines at once.
Project properties let you set the name, log/crash directory, monitor and target size, and the RAM warning thresholds for x86 and x64 per project. Default values defined in the general settings are applied automatically when a new project is created — every project stays independently editable afterwards.
An optional chart shows CPU, RAM and sub-process count of the monitored target process as a time series — the last 2 minutes at a 1-second interval, each series toggled individually via the legend, height freely adjustable via a splitter.
Nothing missed: the trend stays visible even after a brief CPU spike — not just the current value in the status bar.
Three short guides — depending on whether you're setting up GUI Test Studio, using it for the first time, or want to get more out of daily use.
x86/x64 choice, storage location, AI API key, project properties, central defaults.
For new usersFrom a new project to your first automated test run — step by step.
For daily useFine-tune image anchors, hotkeys, debugger attach, pinning the metrics overlay.
Reach out — we're happy to show you how GUI Test Studio fits into your environment.
Download GuiTestStudio