Wendigos Voice Cloning

v2.0.0 - The VoiceBox Rewrite

• Rewrite AI services to use VoiceBox
  • Add support for Gemini, Claude, and Ollama chat services
  • Add support for Elevenlabs STT service
  • Significantly improve real-time response latency using TTS audio streaming
• Rewrite audio networking approach to improve reliability and eliminate lag on game join
• Rewrite local mod functionality
  • Remove local voice cloning model and reliance on main.exe
  • Implement clip collection with speech transcription and smart chatbot clip selection
• Implement optional config sync feature to streamline setup for clients and sharing service API keys
• Add autoclone functionality to allow automatic cloning of player voices during rounds
• Rework dependencies to enable cross-mod compatibility

v1.0.10

• Update Netcode patch to fix v73 error

v1.0.9

• Fixed a bug where masked could become invincible

v1.0.8

• Add support for realtime responses in multiple languages courtesy of joecabezas on github

v1.0.5 - v1.0.7

• Clamp audio so it isnt too loud
• gpt-4o-mini support
• Update tutorial for realtime responses

v1.0.4

• Compatibility patch for LethalLevelLoader

v1.0.3

• Increase volume of masked at the cost of latency

v1.0.1 and v1.0.2

• Quick patches to fix crash

v1.0.0

• Add realtime responses with ChatGPT and Azure
• Allow players with mod disabled to still hear masked talk
• Allow players to modify chatgpt prompt
• Enable speed optimization for Elevenlabs
• Add damaged line type
• and a bunch more stuff

v0.1.9

• Update README to show off the video I made about the mod!

v0.1.8

• Prevent menu popup when using elevenlabs

v0.1.7

• Add support for all languages XTTSv2 supports
• Add support for special characters like arabic

v0.1.6 and prior

• First releases