parakeet tdt 0.6b v2: transcribe.cpp GGUF GGUF conversions of nvidia/parakeet tdt 0.6b v2 for use with transcribe.cpp. Ported from upstream commit 1b149a3, pinned 2026 04 15. Validated against the NeMo reference at transcribe.cpp commit bf0d0b7 on 2026 04 18. Offline English speech to text. A 0.6B parameter Conformer encoder with a TDT/RNNT transducer decoder. Takes a 16 kHz mono WAV and produces a transcript with optional token level timestamps. Not a streaming model; no multilingual capability (see v3 for that). Downloads Quantization Download Size WER (LibriSpeech test clean) : : F32 parakeet tdt 0.6b v2 F32.gguf 2.47 GB 1.68% F16 parakeet tdt 0.6b v2 F16.gguf 1.24 GB 1.68% Q8 0 parakeet tdt 0.6b v2 Q8 0.gguf 730 MB 1.69% Q6 K parakeet tdt 0.6b v2 Q6 K.gguf 608 MB 1.70% Q5 K M parakeet tdt 0.6b v2 Q5 K M.gguf 547 MB 1.70% Q4 K M parakeet tdt 0.6b v2 Q4 K M.gguf 483 MB 1.72% WER measured on the full LibriSpeech test clean split (2620 utterances) with greedy transducer decoding and no external LM. F32 reference baseline: 1.68%. NVIDIA's self reported number on the same split is 1.69%, so the F32 and Q8 0 ports match the upstream reference within rounding. Usage Build transcribe.cp…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy