whisper large v3: transcribe.cpp GGUF GGUF conversions of openai/whisper large v3 for use with transcribe.cpp. Ported from upstream commit 06f233f, pinned 2026 04 25. Validated against the transformers reference at transcribe.cpp commit 5.6.1 on 2026 04 26. OpenAI Whisper large v3 — converted to GGUF for transcribe.cpp. Multilingual transcription, language detection, and speech translation (audio in any supported language → English text). v3 family adds Cantonese (yue) and uses a 128 bin mel input. Encoder decoder transformer; 30 second windows with chunked long form decoding. Downloads Quantization Download Size WER (LibriSpeech test clean) : : F16 whisper large v3 F16.gguf 2.88 GB 1.81% Q8 0 whisper large v3 Q8 0.gguf 1.55 GB 1.82% Q6 K whisper large v3 Q6 K.gguf 1.21 GB 1.83% Q5 K M whisper large v3 Q5 K M.gguf 1.08 GB 1.84% Q4 K M whisper large v3 Q4 K M.gguf 951 MB 1.86% WER measured on the full LibriSpeech test clean split (2620 utterances) with the transcribe.cpp default decode (greedy, suppress tokens, temperature fallback, segment timestamps enabled). OpenAI's self reported number on the same split is 2.01%. We don't know upstream's exact eval config, but the most likely c…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy