POCKET TOM // LOCAL AI // MEMORY // VOICE // WAKE WORD
Choose the POCKET_TOM folder itself. The GGUF model is found automatically.
No Pocket Tom folder selected yet.
Waiting for the POCKET_TOM folder.
Audio Devices
This step only selects and tests the microphone and speaker. It does not change the working AI core.
Audio devices not enabled yet.
Local Voice Mode
Voice recognition and speech generation run locally in this browser. On the first load,
the public Whisper and Kokoro model files are downloaded and then cached by the browser.
Your microphone audio and Pocket Tom's replies are not sent to a speech API.