Parakeet TDT 0.6B v3

Parakeet TDT 0.6B v3 is a compact open speech-to-text model with 600 million parameters (FastConformer encoder with a TDT decoder). Per the model card, it transcribes 25 European languages (including English, Spanish, French, German, Italian, Portuguese, Polish, Ukrainian and Russian), detects the language automatically without extra prompting, adds punctuation and capitalization, and returns word-level and segment-level timestamps. It transcribes audio up to 24 minutes long with full attention on an A100 80 GB, or up to 3 hours with local attention; a streaming script is provided. Hardware: the card says at least 2 GB of RAM is needed to load the model (more RAM allows longer audio) and the model is optimized for NVIDIA GPUs. On the Open ASR Leaderboard (English short-form speech, nine test sets, checked 2026-10-20) it has an average word error rate of 6.51%, with a very high processing speed. Released 2025-08-04. The weights are public without any acceptance step under the CC-BY-4.0 license, which allows commercial use as long as you credit the authors. The authors offer an official demo Space on Hugging Face. Date of check: 2026-10-20.