Voice, network, and model-related issues.

Technical issues

When a technical exception occurs, update to the latest version first, then reproduce once with minimum configuration: turn off proxy/VPN and keep only necessary features before continuing troubleshooting.

API KEY error / network error

Update to the latest version first. If the issue still exists, report it in the after-sales group.

Voice stutters during use

Try either method to recover:

Press Ctrl + M to restart the voice module.
Completely quit and restart the software.

Window is invisible but the process is still running in the background

Use the monitor-switch shortcut to move the window back to the current screen.
If it is still invisible, restart the software and reselect the monitor in settings.

Currently recommended models

Ultimate plan includes Claude, GPT, Gemini, and other models, with stronger overall capability.
Gemini responds faster and is suitable for latency-sensitive scenarios.

Auto-trigger mechanism

After "auto trigger" is enabled, AI answers are called automatically only when the interviewer speaks.
If you ask a question proactively, you still need to press Cmd + G on macOS or the corresponding shortcut manually.

Knowledge-base upload files are skipped

Check whether the file format meets requirements. .txt is recommended.
Confirm that the file content is not empty and does not contain image-only content.

Interview / exam records were not uploaded

Version 2.0.6 and above saves interview and exam records locally. You can find corresponding records in local documents.
If there are no records in local documents either, check whether the session ended normally and contact support with feedback.

Common issues

Troubleshoot frequent exceptions that can recover quickly first.

Contact us

Submit environment information to support when failures continue.