Release ChatGPT CLI v1.8.1 · kardolus/chatgpt-cli

New Features:

Transcription Support:
- Introduced the --transcribe flag to generate transcripts from audio files.
- Supports a variety of audio formats, including .mp3, .mp4, .mpeg, .mpga, .m4a, .wav, and .webm.
- API endpoint used: /v1/audio/transcriptions.
- Appropriate CLI usage will allow users to extract text content from audio files.
Text-to-Speech Support:
- Added the --speak and --output flags for converting text into speech with TTS model gpt-4o-mini-tts.
- Output can be played back directly if afplay is installed (macOS), enhancing accessibility.
- API endpoint used: /v1/audio/speech.

Expanded CLI Capabilities:
- The command-line interface now supports more flexible configuration options using a layered system (default values, config.yaml, environment variables).
- Enhanced debug options for better visibility into API interactions.
Model Capabilities:
- Introduction of a getCapabilities() method to dynamically determine model features such as support for temperature and TTS capabilities.
Modular Endpoints:
- API paths are now more customizable with additions for specific model types, improving integration flexibility.

Addressed issues in API endpoint selection logic which previously led to incorrect routing.
Fixed condition checks related to the new-thread flag for the command-line interface.

Added significant refactoring in the codebase for better readability and maintainability of various modules, including the client, api, and cmd packages.
New mock interfaces and test enhancements using gomock for robust unit testing and improved CI/CD performance.

Update Methods:
- Homebrew Users: Run brew upgrade chatgpt-cli to get the latest update.
- Direct Download: You may download available binaries specific to your OS directly from the repository or website.
Configuration Modification:
- Ensure to update your config.yaml to leverage new features and flags introduced in this release.