speak-gpt

• Added ability to change host if OpenAI is banned in your country.

• Previous version went so tricky and I added a switch that allow you to disable buggy features like function calling or chat commands.

• Added TTS voice selector
• Now you can generate images in any language. "/imagine" command replaced with GPT function calling.
• Little UI fixes
• Added always speak mode

• Added turkish language
• Fixed assistant layout
• Added crash handler

• Removed max tokens property. Now tokens are calculated dynamically. This approach will minimize max tokens error.
• Added Polish language

• Fixed accessibility issues on devices with small screens
• Added Spanish language

• Added automatic detection of TTS language.

• Removed deprecated models
• Minor bugs fixed
• UI improvements

• Added support of multiple languages of voice assistant

Simplified UI

• Optimized for small screens and other UI improvements

• Fixed bug when chat settings are reset after renaming chat.

• Added option to export/import chats
• Now you can set personalized settings for every chat
• Various bugs fixed

• Now you can save generated images
• Various bugs fixed

• Added Whisper speech recognition services.