Using a HomePod Mini as a voice player in Home Assistant (HA) essentially means treating it as a DLNA/AirPlay media player.

Because the built-in Google TTS or cloud services that require internet access often fail in some network environments, I recommend the most stable and best-sounding option available locally: Microsoft Edge TTS. It needs no special network setup, sounds natural, and is completely free.

Below is a detailed tutorial based on Home Assistant version 2025.12.2.


Step 1: Add the HomePod Mini to Home Assistant

Home Assistant has native Apple protocol support and will usually discover the device automatically.

  • Automatic discovery: make sure the HomePod Mini is on the same LAN as HA. Open the HA interface and click Settings > Devices & Services.
  • Add the integration: if you see Apple TV or HomeKit Accessory Protocol in the Discovered card, click Configure.
  • Manual addition: if it is not discovered automatically, click Add Integration in the lower right corner, search for Apple, and select Apple TV, which is the unified integration for Apple devices.
  • Pairing: HA will ask you to enter a verification code on the HomePod. Your phone or iPad, acting as the home hub, will pop up the code. Enter it into HA.
  • Confirm: once successful, you will see an entity of type media_player under Devices, for example media_player.homepod_mini.

Step 2: Install Edge TTS (the best voice solution for local networks)

To make the HomePod speak Chinese, we need a voice engine.

  • Go to Settings > Devices & Services > Add Integration.
  • Search for Microsoft Edge TTS.
  • Note: if it is not in the official store, first make sure HACS is installed and search in HACS.

After installation it provides a tts.microsoft_edge_tts_cloud service.

Set Chinese: in the integration options, it is recommended to set the default voice to zh-CN-XiaoxiaoNeural (Xiaoxiao, the most natural Chinese female voice).


Step 3: Configure the basic TTS options

In the latest HA version, to make voice announcements flow smoothly, we need to quickly confirm a few things in configuration.yaml, or configure them directly in the UI.

Click Developer Tools > Services in the left sidebar.

Enter tts.speak to test whether it works.

Important notes

: for the HomePod to receive audio files, the HA internal URL must be configured correctly.

  • Go to Settings > System > Network.
  • Make sure the Home Assistant URL is set to your LAN IP (for example http://192.168.1.100:8123), and is not a public address.

Step 4: A beginner automation tutorial, the welcome home announcement

Now let us build the simplest automation: when someone arrives home, the HomePod greets them automatically.

  • Go to Settings > Automations & Scenes > Create Automation.
  • Add a trigger: select State.
  • Entity: person.your_name (your user).
  • From: not_home, To: home.

Add an action: select Call Service.

  • Service: tts.speak (note: version 2025 recommends speak rather than the old say).
  • Media Player: choose your media_player.homepod_mini.
  • TTS engine: select Edge TTS.
  • Message: enter Welcome home, thanks for your hard work today!

Save and test: click the three dots in the upper right of the action and select Run to see whether your HomePod speaks.


Advanced: how to make the voice smoother

Because the HomePod has a small delay before it plays audio, it is recommended to configure the automation actions in this order:

  • Action 1: media_player.volume_set (set the volume to 0.5 so it does not startle anyone).
  • Action 2: tts.speak (send the voice message).

Common troubleshooting (pitfall guide):

  • No sound?
  • Check whether AirPlay permission for the HomePod is disabled in the Home app. Set it to Anyone on the Same Network.
  • Confirm whether the HA IP address is fixed. If the HA IP changes, the HomePod cannot find the audio files.

Sound cutting in and out?

  • This is because connections to Microsoft Edge servers occasionally fluctuate. No VPN is needed, but if the network is very poor, configure Media Source caching in HA.

How do I switch to a male voice?

  • In the Edge TTS integration settings, set zh-CN-YunxiNeural as the default. It is a very natural male voice.

With the steps above, you now have a fully localized HomePod voice announcement system with minimal network dependency and excellent sound quality.

If you see the error no response to SETUP rtsp://…, its core meaning is that the handshake failed. You only need to allow UDP ports 10000-65535 through your firewall.

相关新闻

cloud@haiots.com

QQ
微信