Google seems to have unintentionally previewed the next steps for its Gemini AI assistant. Sources at TestingCatalog report that a fresh Troubleshooting mode has appeared in the Gemini model selection menu for certain users.

The new choice appears alongside familiar selections such as Gemini 3.5 Flash and 3.1 Pro, the standard models users already switch between inside the app.

How Gemini Troubleshooting Mode Works in Practice

Rather than delivering lengthy text blocks, Troubleshooting mode leads users through each issue one step at a time by blending written guidance with clickable interactive widgets.

As an example, if a user reports that a car will not start, the system may list frequent causes such as a dead battery and then display tap-to-select symptom choices like clicks or silent to speed up diagnosis. The result is a more organized and directed flow than a regular chat conversation.

Key Differences From Everyday Gemini Queries

This distinction stems from internal tuning. Early testers on Reddit indicate the mode uses a lower temperature setting so it stays tightly focused on the problem and avoids casual conversation.

Its answers prioritize diagnosis and real-world solutions instead of general background. Google has not made any official statement about the feature, so it remains unknown whether this marks a planned release or simply an internal experiment.

At present the Troubleshoot option looks like an accidental activation, implying Google may have enabled it by mistake and could remove it without notice. Further details are expected over the next few weeks.

Users curious about the new Troubleshooting mode will also notice other Gemini updates underway. Google recently launched Gemini Spark, a nonstop AI agent that manages tasks in the background. At the same time free accounts may soon meet stricter weekly usage limits.