speech-to-text

The Speech-to-Text skill leverages ElevenLabs Scribe v2 to convert spoken audio or video content into written text efficiently and accurately. It solves the problem of manually transcribing speech, enabling users to quickly generate text from audio sources such as meetings, interviews, lectures, or multimedia content. By automating transcription, this skill reduces time and effort while maintaining high accuracy, even for complex or multilingual content. It is especially valuable for professionals who need reliable and fast transcriptions for documentation, subtitling, or analysis purposes.

Key features of this skill include support for over 90 languages, word-level timestamps, speaker diarization, and real-time transcription options. Users can specify key terms to improve recognition of specialized vocabulary, and the skill can automatically detect the spoken language with probability scores. It handles a wide variety of audio and video formats, including MP3, WAV, M4A, FLAC, MP4, AVI, MKV, and more. The transcription output provides detailed information, including the text, language code, timestamps for each word, and speaker identification when applicable, making it versatile for both batch and live transcription scenarios.

This skill is particularly useful for content creators, transcription services, educators, and businesses that need to process spoken content into actionable text. Use cases include generating subtitles for videos, transcribing meetings or interviews with multiple speakers, producing written records of lectures or podcasts, and integrating transcriptions into workflows for data analysis or knowledge management. Target users include Python and JavaScript developers, multimedia professionals, and teams working with audio-rich content who require accurate, scalable, and feature-rich transcription capabilities.