This skill provides a standardized implementation for performing speech-to-text transcription using Azure's REST interface. It is specifically designed for 'short audio' scenarios—files up to 60 seconds in length—making it a lightweight alternative to the full Azure Speech SDK. By utilizing standard HTTP requests, it allows developers to integrate high-quality speech recognition into Python applications and Claude Code workflows with minimal dependency overhead. The skill includes patterns for chunked transfer encoding, detailed metadata retrieval, profanity filtering, and asynchronous processing.
主要功能
01Support for audio files up to 60 seconds with simple and detailed response formats
02Chunked transfer encoding implementation for optimized latency
03Standardized error handling for recognition status and API authentication
0431,721 GitHub stars
05Configurable profanity masking and multi-language support (100+ locales)
06SDK-free implementation using Python's requests and aiohttp libraries
使用场景
01Integrating speech recognition into serverless functions where package size is constrained
02Building lightweight voice command interfaces for CLI tools
03Automating transcription of short voice memos or audio snippets