--- title: Wyoming Protocol description: Connect remote voice services to Home Assistant. ha_category: - Voice ha_iot_class: Local Push ha_release: '2023.5' ha_codeowners: - '@balloob' - '@synesthesiam' ha_domain: wyoming ha_integration_type: integration ha_platforms: - binary_sensor - number - select - stt - switch - tts ha_config_flow: true ha_zeroconf: true --- The **Wyoming** {% term integration %} connects external voice services to Home Assistant using a [small protocol](https://github.com/rhasspy/rhasspy3/blob/master/docs/wyoming.md). This enables [Assist](/voice_control/) to use a variety of local [speech-to-text](/integrations/stt/), [text-to-speech](/integrations/tts/), and [wake-word-detection](/integrations/wake_word/) systems, such as: - Whisper {% my supervisor_addon badge addon="core_whisper" %} - Piper {% my supervisor_addon badge addon="core_piper" %} - openWakeWord {% my supervisor_addon badge addon="core_openwakeword" %} ### Installing a local voice pipeline The Wyoming, Piper, and Whisper add-ons are used to create a fully local voice pipeline. To learn how to do this, follow the procedure on [installing a local voice pipeline](/voice_control/voice_remote_local_assistant/). {% include integrations/config_flow.md %} ## Satellites [Remote voice satellites](https://github.com/rhasspy/wyoming-satellite) can be connected to Home Assistant using the Wyoming protocol. These satellites typically run on Raspberry Pi's, and are automatically discovered by Home Assistant through [Zeroconf](/integrations/zeroconf). ### Audio Settings The following settings control audio processing of a satellite's microphone input: - Noise suppression - Level of noise suppression (uses [webrtc]). Audio distortion may occur as the level increases. - Auto gain - Automatically adjusts volume based on ambient noise (uses [webrtc]). The setting value is the target dBFS. - Mic volume - Fixed multiplier applied to microphone audio samples. 2.0 doubles the volume, while 0.5 is halves it. Values above 1.0 may increase noise or cause audio distortion. [webrtc]: https://github.com/rhasspy/webrtc-noise-gain