Step ASR Streaming Transcription
Transcribes audio files into text using the StepFun ASR API with support for real-time streaming and multi-format processing.
The Step ASR skill provides high-performance speech-to-text capabilities within Claude Code by leveraging the StepFun ASR streaming API. It enables users to convert Chinese and English audio into text across various formats including PCM, WAV, MP3, and OGG. With advanced features like terminology correction prompts and real-time streaming output via HTTP SSE, this skill is perfect for developers needing precise transcriptions of technical discussions, meeting recordings, or multimedia content while maintaining control over output formats and metadata.
主な機能
01Real-time streaming output using HTTP SSE for instant feedback
02Comprehensive format support for MP3, WAV, OGG/OPUS, and raw PCM
03Configurable output options including raw text, JSON files, and usage statistics
04Supports high-accuracy transcription for Chinese and English languages
05Custom terminology prompts to improve accuracy for domain-specific jargon
0615 GitHub stars
ユースケース
01Generating automated transcripts and subtitles for audio-based content workflows
02Integrating speech-to-text processing into CLI-based developer productivity pipelines
03Converting technical meeting recordings into text with specific industry terminology correction