diff --git a/docs/USAGE.md b/docs/USAGE.md
index 04864fe..fa09ac9 100644
--- a/docs/USAGE.md
+++ b/docs/USAGE.md
@@ -1773,8 +1773,22 @@ Configuration (optional):
 
 ```toml
 [voice]
-whisper_url = "http://192.168.122.1:8080/inference"
-piper_url = "http://192.168.122.1:5000/"
+whisper_url = "http://192.168.129.9:8080/inference"
+piper_url = "http://192.168.129.9:5100/"
 silence_gap = 1.5
 trigger = ""
 ```
+
+Piper TTS accepts POST with JSON body `{"text": "..."}` and returns
+22050 Hz 16-bit PCM mono WAV. Default voice: `en_US-lessac-medium`.
+
+Available voices:
+
+| Region | Voices |
+|--------|--------|
+| en_US | lessac-medium/high, amy-medium, ryan-medium/high, joe-medium, john-medium, kristin-medium, danny-low, ljspeech-high |
+| en_GB | alba-medium, cori-medium/high, jenny_dioco-medium, northern_english_male-medium, southern_english_female-low |
+| fr_FR | siwis-low/medium, gilles-low, tom-medium, mls-medium, mls_1840-low, upmc-medium |
+
+To switch the active voice, set `piper_tts_voice` (e.g.
+`fr_FR-siwis-medium`) and redeploy the TTS service.