Version 1.2.0

Changelog

• Fixed incompatibility of speech_recognition library with OpenAI's new API versions.
• Updated to GPT-4o
• Saying case to continue talking is not needed anymore, users can continue to talk in a natural manner.
• Updated readme
• Improved code readability
• Reduced code complexity

Committed 01.07.2024
Full Changelog: v1.1.0...v1.2.0

Version 1.1.0

Changelog:

• Switched to the newest OpenAI API release, bringing a selection of more natural voice models.
• New voice models allow users to receive answers in many languages with accurate pronunciation.
• Switched to Google's speech-to-text model, allowing increased wait time for the call phrase and less cost.
• Checks if required plugins are installed at each startup, and installs them if necessary.
• Optimised and more readable code.

Ability to speak to Case in other languages will be added later, hopefully before EoY
Committed 24.12.23
Full Changelog: v1.0.2...v1.1.0

Version 1.0.2

Changelog:

• OpenAI API key is now inputted from the console, at startup. Much more user-friendly.
• Better call phrase recognition.
• Coherent website content.
• More readable code.

Full Changelog: v1.0.1...v1.0.2

Committed 22.11.23

Version 1.0.1

Changelog:

• Critical fix to requirements

Full Changelog: v1.0.0...v1.0.1

Committed 21.11.2023

Version 1.0.0

CaseAI

Voice-to-voice artificial intelligence assistant software that can be run in a Python IDE with a web interface. Powered by OpenAI's GPT 3.5 Turbo and Whisper speech recognition API's.

Requirements:

• Python 3
• OpenAI API key by the user.
• Python IDE to run the application.

How to use:

1. Get an OpenAI API key. It is free to open an OpenAI account and gives you plenty of credits for a few months. Here is how to create an API key, given you have an account.
2. Run the application on a Python IDE and access the web interface via the local IP given in the console.
3. Press "Start Talking". You can now start talking with case by saying the activation phrase "Hey Case"

Committed 25.09.2023