Voice Command Mod

Say something you shouldn't. Watch something spawn. Shown best by Xeroh.

Voice Command Mod listens to your microphone while you play Minecraft. When it detects a trigger word in your speech, it randomly selects a mob and spawns it near you. No Python, no external programs, no setup beyond dropping the JAR in your mods folder.

Built for streamers, but works for anyone who wants a reason to watch their mouth.

How It Works

The mod uses Vosk for offline speech recognition. Everything runs locally on your computer. A small voice model(around 40MB) is extracted to your .minecraft folder on first launch automatically. No internet connection required for voice detection, and no audio data is ever sent anywhere.

Requirements

Minecraft 1.20.1
Forge 47.3.0 or higher
wonderland.jar (required to spawn preset mobs, HOWEVER, this can be changed in the config to any mob id you want)
A working microphone set as your default recording device in Windows

Features

Voice Triggered Mob Spawns say a trigger word(changeable in config) and a mob from that word's pool spawns. Each word group has a list of mobs and one is chosen at random every time.
Substring Detection the mod catches trigger words inside longer words. Without this, the mod was a bit boring. This was added for some extra chaos.
Cursed Lottery every 10th word activates the Cursed Lottery. Instead of one mob, 3 to 5 random mobs from the entire pool spawn rapidly.
Shame Title every time a word triggers a message appears in chat showing your total swear count and the word you said.
Clean Streak Tracker after each trigger the time since your last trigger word is displayed above your hotbar so you and your viewers can see how long you managed to behave.
Fully Configurable Word and Mob Lists (As of version 2.1.1 there is an in game config menu) On first launch the mod creates a plain text config file at

.minecraft/config/voicecommand_words.txt

Open it in Notepad to add words, remove words, change which mobs each word spawns, and control which mobs spawn on top of you verus in front of you. No JSON, no special syntax, just plain text. Example of adding a new trigger word:

chicken = the_wonderland:carrier_aggro, the_wonderland:eyeless

In Game Settings additional settings are avaiable through the Forge mod config screen under Mods -> Voice Command Mod -> Config:
- Enable or disable mod entirely
- Trigger cooldown in seconds
- Lottery frequency
- Toggle swear count display
- Toggle streak display

First Launch

On your very first launch with the mod installed, the Vosk voice model will be automatically extracted to your .minecraft folder. This happens in the background and takes a few seconds. You will see this message in your game log when it is ready:

[VoiceMod] Microphone started - listening for trigger words!

For Streamers

This mod was designed with streaming in mind. The shame title appears in chat so viewers can see them in stream. The streak tracker gives viewers something to watch build up. The Cursed Lottery creates natural moments of tension as the swear count climbs toward the next jackpot.

The word and mob config file makes it easy to customise the mod for your specific modpack without needing to recompile anything.

Troubleshooting

Nothing happens when I speak

Make sure your microphone is set as default recording device in Windows Sound Settings
Check the game log for the line "[VoiceMod] Microphone started"
Make sure you are fully loaded into a world, not just on the title screen

Mobs are not spawning

Make sure the wonderland.jar is installed(unless changed in configs)
Check the config file to confirm the mob IDs are correct
Check the game log for any lines starting with [VoiceMod] ERROR
Make sure you have permissions to spawn mobs (only in version 2.0.0)

I want to use mobs from a different mod

Open .minecraft/config/voicecommand_words.txt
Replace the mob IDs with any valid entity ID from your modpack
The format is modid:entityname, the same as the /summon command

Privacy

This mod uses the Vosk speech recognition library which runs entirely offline. No audio is recorded to disk. No audio or transciption data is ever sent to any server. Everything stays on your machine.