Add support for Cloud Based LLMs and Cloud based inference

[connorakey]

This pull request adds support for using cloud-based, OpenAI-compatible APIs for both text cleaning (LLM) and audio transcription, in addition to the existing local providers (Ollama, LM Studio, and Whisper). It introduces new configuration options, updates the backend logic to support cloud providers, and documents these changes for users. The most important changes are grouped below:

Backend feature enhancements:

• Added support for using any OpenAI-compatible API (e.g., OpenAI, Together AI, OpenRouter, Azure OpenAI, Groq) as a provider for both LLM text cleaning and audio transcription. This includes new environment variables (CLOUD_API_BASE_URL, CLOUD_API_KEY, TRANSCRIPTION_PROVIDER, TRANSCRIPTION_MODEL_NAME) and logic to handle cloud API requests and responses in both text_cleaner.py and transcription.py [1] [2] [3] [4] [5] [6].

• Implemented a new _transcribe_cloud function in transcription.py to support audio transcription via cloud APIs, including authentication and model selection.

Configuration and dependency updates:

• Updated .env.example and documentation to include new configuration options for cloud providers and transcription settings, clarifying usage for different deployment scenarios [1] [2] [3].

• Added the requests library as a dependency in pyproject.toml to support HTTP requests to cloud APIs.

Documentation improvements:

• Expanded docs/API.md to provide detailed instructions and examples for configuring and using cloud providers for both LLM and transcription, including provider-specific examples and a discussion of benefits and considerations [1] [2].