Qwen3 ASR 1.7B hf
Qwen3-ASR-1.7B-hf is an Apache-2.0 automatic speech recognition model from the Qwen3-ASR family. It is available through Hugging Face and supported natively in Transformers starting from v5.13.0 (per the model card). The model card describes it as supporting language identification and ASR for 30 languages and 22 Chinese dialects, with offline and streaming inference, and it lists support for speech, singing voice, and songs with background music.
At a glance
- Parameters
- 2.04B
- Architecture
- qwen3_asr
- License
- apache-2.0
- Disk space
- 3.8 GB
- Software
- Transformers
Will it run on your machine?
Save your machine to see a personalized rating and its reasoning.
Add my hardwareBest for
- Automatic speech recognition for the languages and dialects listed in the model card.
- Language identification alongside transcription.
- Offline or streaming transcription, as described in the model card.
- Transcription of speech, singing voice, or songs with background music, as listed in the model card.
- Long audio handling, as described in the model card.
Tradeoffs
- The supplied facts do not provide license terms beyond the Apache-2.0 license.
- The supplied facts do not provide hardware requirements or performance benchmarks.
- The supplied facts do not confirm compatibility beyond the stated Transformers version requirement.
- The supplied facts do not list supported languages beyond those in the model card.
- The supplied facts do not state limitations for specific audio conditions or domains.
Ways to run it
Transformers · See official guide
Repository-specific command in publisher documentation; confirm dependencies and hardware in the source.
from transformers import AutoProcessor, AutoModelForMultimodalLM
model_id = "Qwen/Qwen3-ASR-1.7B-hf"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForMultimodalLM.from_pretrained(model_id, device_map="auto")
print(f"Model loaded on {model.device} with dtype {model.dtype}")
inputs = processor.apply_transcription_request(
audio="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen3-ASR-Repo/asr_en.wav",
).to(model.device, model.dtype)
output_ids = model.generate(**inputs, max_new_tokens=256)
generated_ids = output_ids[:, inputs["input_ids"].shape[1]:]
# Raw output includes language tag and <asr_text> marker
raw = processor.decode(generated_ids)[0]
print(f"Raw: {raw}")
# Parsed output: dict with "language" and "transcription"
parsed = processor.decode(generated_ids, return_format="parsed")[0]
print(f"Parsed: {parsed}")
# Extract only the transcription text
transcription = processor.decode(generated_ids, return_format="transcription_only")[0]
print(f"Transcription: {transcription}")
"""
Raw: language English<asr_text>Mr. Quilter is the apostle of the middle classes, and we are glad to welcome his gospel.
Parsed: {'language': 'English', 'transcription': 'Mr. Quilter is the apostle of the middle classes, and we are glad to welcome his gospel.'}
Transcription: Mr. Quilter is the apostle of the middle classes, and we are glad to welcome his gospel.
"""Explore its uses
Keep exploring
Voxtral Mini · 4B Realtime
Voxtral Mini 4B Realtime 2602 is a multilingual, realtime automatic speech recognition model from Mistral AI, released under the Apache-2.0 license. It features a natively streaming architecture with a causal audio encoder and sliding window attention, enabling configurable transcription delays from 240ms to 2.4s (and reportedly 80ms to 2.4s). The model has approximately 4B parameters (≈3.4B language model and ≈970M audio encoder) and is described as optimized for on-device deployment. It supports 13 languages and is presented as achieving accuracy comparable to offline systems with less than 500ms delay, specifically matching leading offline open-source models and realtime APIs at a 480ms delay. The model card recommends a temperature of 0.0, using websockets for audio streaming, and a delay of 480ms as a balance between performance and latency.
YuE2 · 3B
Create a full song from style and lyrics with a publisher Python package on a Linux NVIDIA GPU; outputs include a playable FLAC.
granite speech 5.0 470m turboctc
Granite Speech 5.0 470M TurboCTC is a compact 470 million parameter English automatic speech recognition model released by IBM under the Apache 2.0 license. It uses a conformer acoustic encoder with block self-attention, self-conditioning, and temporal downsampling, with an output layer of 16,384 BPE units. The model was trained on approximately 60,000 hours of English audio from public corpora using Connectionist Temporal Classification (CTC), and inference is non-autoregressive with greedy decoding. It is designed for low-latency and high-throughput English speech-to-text transcription, and is described as suitable for deployment on laptops, smartphones, and other edge devices. Native support is available in transformers>=5.16.0, with additional usage options for Apple Silicon via mlx-audio and local GGUF inference via transcribe.cpp.