What problem does it solve? Converting spoken audio into accurate written text manually is slow and error-prone, especially for long recordings, multi-speaker meetings, or content in multiple languages. This Skill provides ready-to-use instructions for transcribing audio and video files with ElevenLabs Scribe v2, including speaker identification and word-level timestamps. ## Core Features & Use Cases - Batch Transcription: Convert audio and video files (MP3, WAV, MP4, and more) to text across 90+ languages with automatic language detection. - Speaker Diarization & Timestamps: Identify who said what with speaker labels and word-level timing, ideal for subtitles and meeting notes. - Real-Time Streaming: Transcribe live microphone input or audio streams with ~150ms latency using the scribe_v2_realtime model and VAD-based commit strategies. - Use Case: Transcribe a recorded team meeting with diarization enabled to produce a speaker-labeled transcript, then generate SRT subtitles from the word-level timestamps. ## Quick Start Ask the agent to transcribe an audio file such as meeting.mp3 into text with speaker labels using the ElevenLabs speech-to-text API.
