Microphone Text Input

What's New

• use FreshSupaSulley/Hhisper-jni instead of GiviMAD/whisper-jni, see https://github.com/GiviMAD/whisper-jni/issues/18
• disable custom dynamic library loading
• support vulkan on Windows x64 (Advanced Settings)
• only support Windows x64, Mac (both AMD x64 and ARM), and Linux (both AMD x64 and ARM), see FreshSupaSulley/whisper-jni

Dependencies

What's New

• Support for Fabric & NeoForge

Dependencies

Whats New

• Support for Fabric & NeoForge

Whats New

• Compatibility with 1.21.4, 1.21.5 and 1.21.6.
• Add missing translations for file chooser dialog.

Installation

1. Install Minecraft 1.21.4, 1.21.5 or 1.21.6, fabric loader >=0.16.0, fabric-api and MidnightLib 1.7.2-fabric or an updated version.
2. Install the mod and download the GGML Whisper model.
3. Configure the mod according to your needs in the configuration menu.
4. Use the keybinding you set (default is V) to start recording and speech recognition.
5. If the mode is set to AUTO_SEND, the recognized text will be automatically sent as a chat message. If set to RELEASE_KEY_TO_SEND or RELEASE_KEY_TO_INPUT, follow the corresponding key - release actions.

For more info, see README: 2.x

Whats New

• OpenAI Whisper model support.
• Support custom dynamic link libraries, the users can compile custom libraries for CUDA support.

Installation

1. Install minecraft 1.21.5, fabric loader >=0.16.0, fabric-api and MidnightLib 1.7.2-fabric.
2. Install the mod and download the GGML Whisper model.
3. Configure the mod according to your needs in the configuration menu.
4. Use the keybinding you set (default is V) to start recording and speech recognition.
5. If the mode is set to AUTO_SEND, the recognized text will be automatically sent as a chat message. If set to RELEASE_KEY_TO_SEND or RELEASE_KEY_TO_INPUT, follow the corresponding key - release actions.

Publish by GitHub Action. For more info, see README: 2.x

• optimize UI

For more info, see GitHub Releases 2.0.0

• add whisper log level config

• all UN languages are supported

New Features

Use OpenAI whisper model instead of VOSK.

Usage

1. Install the mod and download the GGML Whisper model.
2. Configure the mod according to your needs in the configuration menu.
3. Use the keybinding you set (default is V) to start recording and speech recognition.
4. If the mode is set to AUTO_SEND, the recognized text will be automatically sent as a chat message. If set to RELEASE_KEY_TO_SEND or RELEASE_KEY_TO_INPUT, follow the corresponding key - release actions.

More Info

https://github.com/Jaffe2718/Fabric-Microphone-Text-Input/tree/2.x

• update to mc 1.21.4

• Translations have been optimized

• Optimized UI
• Translations have been optimized
• The server-side entry point has been removed and is now entirely a client-side mod
• Access modmenu to check for updates API

• Minecraft 1.21/1.21.1 is supported;
• Not sure about 1.20.5/1.20.6 compatibility, you can try it out;