How to change voice in the robot vacuum cleaner Xiaomi Roborock: official and informal methods

Xiaomi Roborock’s robot vacuum cleaners have become an integral part of smart homes, but the standard voice assistant can get bored or seem too monotonous. Fortunately, the manufacturer and enthusiasts offer several ways to change voice prompts, from official packages in branded apps to custom solutions for advanced users.

In this article, we will discuss all the current methods of changing voice for Roborock models. S7, S8, Q7 Max and other popular lines.You will learn how to install alternative voice packets via the Mi Home or Roborock App, where to download unofficial voiceovers (including celebrity voices), and how to manually replace audio files through the Mi Home or Roborock App. FTP. We will pay special attention to the nuances for Chinese and global versions of firmware - they differ in the available options.

Let me warn you right away: not all methods are equally safe. Official methods do not require technical skills, but are limited to standard packages. Informal solutions give more freedom, but can lead to loss of warranty or failure of the vacuum cleaner if installed improperly. Choose an approach based on your goals and willingness to experiment.

Official ways: changing voice through Xiaomi apps

The easiest and safest method is to use the built-in features of branded apps. Xiaomi and Roborock periodically add new voice packets that can be installed in a few clicks. Consider both official apps.

In Mi Home (for Chinese versions of vacuum cleaners), the path to voice settings looks like this:

1. Open the app and select your Roborock from the device list.
2. Go to Settings → Voice prompts.
3. In the Voice Package section, tap on the current version (usually Standard or Female/Male).
4. Download and use one of the available packages (the list depends on the region and model).

The Roborock App (for global versions) has a similar algorithm, but the menu can be called Voice Pack or Voice Assistant. Importantly, some packages (such as voice actors or cartoon variants) are only available to users in China. To see them, you need to change the account region - but this can lead to loss of communication with other Xiaomi devices.

Official packages usually include:

• 🗣️ Standard male/female voices with no accents.
• 🎭 Voices of actors – for example, voice acting by Zhang Ziya (only available in Chinese version of Mi Home).
• 🤖 Robotic – Synthetic voice with electronic notes.
• 👶 Children's - high and playful tone (suitable for families with children).

⚠️ Note: When you change voice through Mi Home, the vacuum cleaner can automatically update the firmware to the Chinese version, which will result in the loss of support for some functions (such as zoning) in global applications. → About the device.

Informal methods: custom voice packets

If there are not enough formal options, you can install informal voice packs, enthusiasts create voiceovers in different languages, with humor or to specific queries (for example, the voice of Darth Vader or Terminator), the main risk is incompatibility with your model of vacuum cleaner or firmware.

The most popular sources of custom packages:

• 🌐 Forums 4PDA — Roborock topics often include archives with voices and instructions.
• 📦 GitHub – open-pack repositories (e.g. roborock-voice-packs).
• 💬 Telegram channels – Xiaomi smart devices communities share fresh builds.

The installation process usually includes:

1. Download the archive with a voice packet (.zip or.tar format).
2. Connection to the vacuum cleaner FTP (It is necessary to activate the developer mode).
3. Replace files in the folder /miio/voice or /roborock/voice (path depends on the model).
4. Reboot the device.

Download the package for your Roborock model|

Check the firmware version (Chinese/global)|

Create a backup of original audio files|

Ensure a stable Wi-Fi connection|

Have access to FTP-Customer (e.g. FileZilla)-->

Example of a path for the Roborock model S7:

ftp://192.168.1.100/miio/voice/


(where 192.168.1.100 is the IP- address of the vacuum cleaner in your network)

⚠️ Attention: Some custom packs contain audio files with increased volume or distortion. Before installing, check other users' feedback on a particular build. For example, Yoda's voice from Star Wars may sound too quiet for large spaces.

Change voice through developer mode (for advanced)

If you're ready to experiment, you can activate the developer mode on the vacuum cleaner and manually replace the audio files, a method that works for Linux-based models (most of Roborock after 2019).

• 🖥️ Computer with FTP-client (e.g. FileZilla or WinSCP).
• 📡 Access to the local network (vacuum cleaner and PC must be in the same subnet).
• 🔧 Knowledge of the basics of working with the console (to check access rights).

Step-by-step:

1. Activate Developer Mode: Press the Home and Spot buttons on the vacuum cleaner for 5 seconds. You will hear an audible beep and the indicator will flash orange.

• IP-Address: check in the router or via the command ping roborock.local in the terminal.
• Login/password: usually root and 123456 (or blank password for new models).
• Go to the folder with the audio files (see the table below) and replace them with your own (.wav format, 16 kHz bitrate).
• Reboot the vacuum cleaner with the reboot command through FTP or long press of the power button.

To create your own audio files, use programs like Audacity with settings:

• 🎵 Format: WAV PCM (compressively).
• 📊 Bitrate: 16 bits, sample rate 16 kHz.
• 🔊 Volume: normalize to -3 dB to avoid distortions.

Problems after voice change and their solution

Even if you follow the instructions, you can have problems, and let's look at the typical situations and how to fix them.

The vacuum cleaner does not make sounds after replacing the files:

• 🔍 Check the file format – it should be in WAV, not MP3.
• 📁 Make sure the files are in the correct folder (see table above).
• 🔄 Reboot the vacuum cleaner with a long press of the power button (10 seconds).

The voice is interrupted or sounds with interference:

• 🎛️ Check the bitrate - it should be exactly 16 kHz.
• 🔊 Reduce the volume of the source file in the editor.
• 📶 Make sure that there is no loss of communication with the vacuum cleaner during the download of files.

The vacuum cleaner got stuck after changing his voice:

• 🔋 Turn it off by holding the power button for 15 seconds.
• 🔄 If it does not help - reset to factory settings (reset button on the case).
• 📥 Reflash the device through Mi Home (Update section).

💡

If after changing the voice, the vacuum cleaner becomes slower to respond to commands, try to remove unused audio files from the folder. /miio/voice/. Sometimes extra files take up memory and slow down the processing of commands.

For diagnostics, it's useful to look at the vacuum logs. SSH (port 22, login/password as for FTP) and execute the command:

cat /var/log/miio.log | grep "voice"

This will show errors related to sound reproduction.

Comparison of methods: which way to choose?

The choice of method depends on your goals, the vacuum cleaner model and your risk-awareness, and below is a comparative table with the pros and cons of each approach.

If your goal is to simply change your standard voice to another official voice, the first method is enough. You need custom packs to set up the voice of Morgan Freeman or Golum of the Lord of the Rings, and if you want to record the voice of a loved one or a pet, you have to master manual file replacement.

💡

For Roborock models on global firmware, the safest way is to use the Roborock App and official packages. Custom solutions are better tested on Chinese versions, where it is easier to roll back to factory settings.

Where to download voice packs for Roborock?

Official packages are installed directly from apps, but where to look for unofficial?

• 🌍 Official Roborock Forum (community.roborock.com) – sometimes put out beta versions of voices.
• 📥 4PDA — Model-specific topics (e.g., Roborock discussion) S7).
• 🐙 GitHub – repositories like roborock-voice-packs or xiaomi-voice-mods.
• 💬 Telegram channels: @xiaomi_robots — News and fashions for robot vacuum cleaners. @roborock_mods — custom firmware and voice packs.
• 🎵 Specialized sites: VoiceMods (for smart devices), Xiaomi Voice Packs (archive of voices for different gadgets).

Before downloading, check:

1. Compatibility with your model and firmware version.
2. Reviews from other users (especially about job stability).
3. Installation instructions – some packages require additional manipulation (e.g., editing configuration files).

Example of file name for Roborock model S6:

roborock_s6_voice_pack_russian_male_v2.zip

Here:

• roborock_s6 — model.
• russian_male — tongue.
• v2 — packet.

⚠️ Note: Some packages from torrent trackers or little-known sites may contain malicious code. Before installing, check the files with antivirus and compare their hash amounts with the specified authors.

Creating your own voice package for Roborock

If you don't find the right voice, you can write down your own.

• 🎤 A microphone with good recording quality (e.g. Blue Yeti or even a smartphone with a Lavalier microphone).
• 💻 Software for audio editing (Audacity, Adobe Audition).
• 📁 Template of sound files for your model (can be extracted from the original package).

Step-by-step:

1. Write down the phrases: Use a neutral tone, no echoes or noises. Examples of phrases: "Bringing clean", "charging completed", "Error: stuck" Follow the official specification (pause length, intonation).
2. Process the sound: Remove background noise (effects) → Noise suppression at Audacity. -3 dB. Save it in format. WAV PCM, 16 bit, 16 keto.
3. File Names must match the original file names (e.g. start.wav, charge_complete.wav). Archive in ZIP compressively.

Set it on the vacuum cleaner:

• Upload the archive to the folder. /miio/voice/ through FTP.
• Reset the device.

Example of a folder structure with voice files for Roborock S7:

voice/




├── start.wav # Start cleaning




├── pause.wav #pause




├── complete.wav # Cleanup complete




├── error.wav # Error




├── low_battery.wav # Low charge




└── charging.wav # Charging

If you want to share your package with the community, upload it to GitHub or forums with the following:

• Vacuum models.
• The version of the firmware that was tested on.
• A list of substituted phrases.

💡

To record voices with emotions (such as cheerful or angry), use text-to-speech services like ElevenLabs or Murph.ai, which allow you to generate realistic speech through text with the desired intonation.

FAQ: Frequent Questions About Robot Change