Scribe

Scribe is a Speech-to-Text model by ElevenLabs. With support for 99 languages, it excels at handling unpredictable real-world audio, delivering transcriptions with word-level timestamps, speaker diarization, and audio event tagging.