The ONLY Free ASR To Transcribe Speech-to-Text in 2025
AI Summary
Nvidia has launched its new automatic speech recognition model called Parakeet TDT, featuring 600 million parameters. The model is designed to provide high-quality English transcription, including punctuation, capitalization, and timestamping, making it ideal for converting audio clips into subtitles. Parakeet TDT performs exceptionally well on the Hugging Face ASR leaderboard, showcasing a low word error rate compared to other models. The video demonstrates its accuracy in transcribing various audio samples and highlights its ability to handle accents effectively. The model is open source and available for commercial use under CC BY 4.0.