FAQ
Answers to common questions about the Mockingbird technical interview assistant: privacy, hardware requirements, Invisible Mode and more.
Do I need an NVIDIA GPU?
No. On CPU recognition works, but slower. A GPU with CUDA acceleratesfaster-whisper and reduces answer latency.
Does my data go to the cloud?
Only the LLM request leaves your machine. Speech recognition, your résumé and the knowledge base are processed locally on your computer.
What is Invisible Mode?
The assistant window is excluded from screen capture: during sharing and streaming it is invisible both in recordings and to viewers.
Which platforms are supported?
Windows 10/11 and Linux (x86_64).
How much does Mockingbird cost?
Free, the source code is open.
The app doesn't see my GPU
Check your CUDA drivers and that the build ships with GPU support. The toolbar indicator should show GPU; otherwise CPU is used.
The answer takes too long
Typical latency is 2–5 s from the end of the question to the answer on GPU. If it's longer — make sure the model is loaded and the GPU is active.
How do I update?
Via the releases pageor through the new-version notification on startup.