Qwen3 ASR 0.6B: transcribe.cpp GGUF GGUF conversions of Qwen/Qwen3 ASR 0.6B for use with transcribe.cpp. Ported from upstream commit 5eb144179a02acc5e5ba31e748d22b0cf3e303b0, pinned 2026 04 19. Validated against the qwen asr 0.0.6 reference at transcribe.cpp commit 3f61df7 on 2026 04 20. Offline multilingual speech to text. An 18 layer bidirectional audio encoder feeds a 28 layer Qwen3 causal LM with audio token injection (fused audio+text sequence, no cross attention). Auto detects the audio's language across 30 languages and emits the transcript in that language. Takes a 16 kHz mono WAV; explicit language hints are not supported at this time. Downloads Quantization Download Size WER (LibriSpeech test clean) : : BF16 Qwen3 ASR 0.6B BF16.gguf 1.46 GB 2.11% F16 Qwen3 ASR 0.6B F16.gguf 1.47 GB 2.12% Q8 0 Qwen3 ASR 0.6B Q8 0.gguf 811 MB 2.11% Q6 K Qwen3 ASR 0.6B Q6 K.gguf 763 MB 2.10% Q5 K M Qwen3 ASR 0.6B Q5 K M.gguf 707 MB 2.21% Q4 K M Qwen3 ASR 0.6B Q4 K M.gguf 654 MB 2.26% WER measured on the full LibriSpeech test clean split (2620 English utterances) with the Whisper style English text normalizer and jiwer 3.x, on the metal backend of an Apple M4. Qwen3 ASR is a multilingual mode…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy