Local Audio

よむ can play audio from any Yomitan-compatible audio source. There are two good ways to set it up:

What you want	Best choice
The easiest setup	Use the hosted Ultimate Yomitan Audio URL
Audio files stored on your own computer	Download and run the local audio server

Short version: use the hosted option if you want the least fuss. Use the local server only if you are okay keeping a small helper app running on your computer.

Easiest: Hosted Audio

Ultimate Yomitan Audio Source gives you a personal audio URL after you subscribe through Patreon and authenticate. That URL already works with よむ, so you do not need to download audio files or run anything on your computer.

Add it to よむ:

1. Open よむ settings with the floating よむ button or Alt+Shift+J.
2. Go to Audio.
3. Press Add audio source.
4. Set Type to Custom URL.
5. Paste the personal URL you were given.
6. Save, look up a word, and press the speaker button.

Local Audio: What You Need

Local audio means よむ asks a helper app on your computer for the sound file.

You need:

1. A computer that stays awake while you study.
2. The audio files.
3. The local audio server download.

The server download is here:

Yomichan/Yomitan Audio Server releases

Do not use the green Code button on GitHub. That downloads developer source code. Normal users want the latest file from the Releases page.

Step 1: Download the Server

Open the release page and download the file that matches your computer:

Computer	File to download
Windows	the file ending in windows-x86_64.zip
Apple Silicon Mac	the file ending in macos-aarch64.tar.gz
Intel Mac	the file ending in macos-x86_64.tar.gz
Linux	the file ending in linux-x86_64.tar.gz

Unzip or open the download. Put the extracted folder somewhere easy to find, such as your Desktop, and rename it to yomu-audio.

Inside that folder you should see a server file named either:

yomichan_audio_server

or, on Windows:

yomichan_audio_server.exe

Step 2: Add the Audio Files

Download the audio files:

nyaa.si/view/1957972

Create a folder named audio inside your yomu-audio folder. Put the downloaded audio source folders inside it.

The folder should look like this:

yomu-audio/
├── audio/
│   ├── daijisen/media/
│   ├── jpod/media/
│   ├── nhk16/media/
│   ├── shinmeikai8/media/
│   ├── forvo_jp/
│   └── forvo_zh/
└── yomichan_audio_server

On Windows, the server file will usually be yomichan_audio_server.exe.

Keep the audio folder names the same. For example, do not rename forvo_jp or nhk16.

Step 3: Start the Server

The server must stay open while you use local audio. If you close the Terminal or PowerShell window, local audio stops until you start it again.

Use port 9393. That avoids a common conflict with other local apps.

Windows

1. Open the yomu-audio folder in File Explorer.
2. Right-click an empty space in the folder.
3. Choose Open in Terminal or Open PowerShell window here.
4. Paste this command and press Enter:

.\yomichan_audio_server.exe --port 9393 --audio .\audio --log full

macOS or Linux

1. Open Terminal.
2. Type cd , including the space after cd.
3. Drag your yomu-audio folder into Terminal.
4. Press Enter.
5. Paste these commands and press Enter:

chmod +x ./yomichan_audio_server
./yomichan_audio_server --port 9393 --audio ./audio --log full

If macOS blocks the app, open System Settings > Privacy & Security and allow it, or Control-click the server file in Finder and choose Open.

Step 4: Check That It Works

Leave the server window open. Open this test link in your browser:

http://localhost:9393/?term=猫&reading=ねこ

If it works, you will see text containing audioSources.

If the browser says the page cannot be reached, the server is not running, the window was closed, or the command used a different port.

Step 5: Add It to よむ

1. Open a page where よむ is running.
2. Open settings with the floating よむ button or Alt+Shift+J.
3. Open Audio.
4. Turn on Enable audio playback for terms.
5. Press Add audio source.
6. Set Type to Custom URL.
7. Paste this exact URL:

http://localhost:9393/?term={term}&reading={reading}

8. Move the local audio source above the built-in sources if you want local audio tried first.
9. Save settings.
10. Look up a word and press the speaker button.

Leave {term} and {reading} exactly as written. よむ replaces those placeholders for each word you look up.

JPDB and browser text-to-speech rows are fallback-only by default, so random audio still prefers recorded clips first. In Settings > Audio, change Text-to-speech handling to Follow source order / random if you want TTS rows to follow your source order or random audio setting.

Using an iPad or Another Device

localhost means "this device."

That means:

• On your computer, localhost:9393 means the computer running the audio server.
• On your iPad, localhost:9393 means the iPad, not your computer.

To use your computer's audio server from an iPad, phone, or second computer, use Tailscale.

Basic setup:

1. Install Tailscale on the computer running the audio server.
2. Install Tailscale on the iPad, phone, or other computer.
3. Sign in with the same Tailscale account on every device.
4. Keep Tailscale connected.
5. Leave the audio-server computer awake.

On the computer running the audio server, run:

tailscale serve --bg --https=443 http://127.0.0.1:9393
tailscale serve status

tailscale serve status prints a private Tailscale URL that looks like this:

https://desktop.your-tailnet.ts.net

Use that URL in よむ on the other device:

https://desktop.your-tailnet.ts.net/?term={term}&reading={reading}

Tailscale Serve keeps the server private to your own Tailscale account. You do not need Tailscale Funnel.

If Audio Does Not Play

• Make sure the server window is still open.
• Make sure the URL in よむ uses 9393.
• Make sure the audio folders are inside yomu-audio/audio/.
• Make sure the audio folder names were not changed.
• If the browser test does not load, start the server again.
• If iPad playback fails, use the Tailscale URL, not localhost.
• If this setup feels like too much, use the hosted audio option at the top of this page. It is much easier.