Description
Qwen3-ASR-1.7B-hf is part of the Qwen3-ASR family, which includes models designed for automatic speech recognition (ASR) and language identification across 52 languages and dialects. Leveraging large-scale speech training data, these models are built on the foundation model Qwen3-Omni, known for its strong audio understanding capabilities. The 1.7B version achieves state-of-the-art performance among open-source ASR models and competes with leading proprietary commercial APIs.
Read More
Key features of Qwen3-ASR-1.7B-hf include support for language identification and speech recognition for 30 languages and 22 Chinese dialects, including various English accents. The model is designed to handle complex acoustic environments with high-quality and fast recognition. It supports streaming and offline unified inference, making it versatile for different audio processing needs.
The Qwen3-ForcedAligner-0.6B model complements the ASR capabilities by providing forced alignment and timestamp prediction for speech segments up to five minutes long in 11 languages. This model surpasses end-to-end forced-alignment models in accuracy, offering precise word-level timestamping.
Qwen3-ASR models are natively supported in the Transformers library, starting from version 5.13.0, allowing for easy integration and use. Users can force transcription languages, pass context for domain-specific vocabulary, and perform batch inference for multiple audio files. The models also support training and fine-tuning, enabling customization for specific applications.
Overall, Qwen3-ASR-1.7B-hf is ideal for developers and researchers looking for a robust, open-source solution for multilingual speech recognition and language identification.
- CategoryAI Models & LLMs
- Pricing ModelOpen source
- DeploymentCloud
- PlatformsWeb
Qwen3-ASR-1.7B-hf Highlights
Multilingual ASR support
Language identification
Streaming and offline inference
Forced alignment with timestamps
High throughput and concurrency
Robust in complex environments
Native support in Transformers
Batch inference capabilities
Getting Started with Qwen3-ASR-1.7B-hf
Access model: Visit Hugging Face
Authenticate: Set up API access
Set up environment: Install dependencies
Integrate via API: Use Transformers library
Optimize: Fine-tune for specific needs
Qwen3-ASR-1.7B-hf's Use Cases
- Multilingual Transcription
- Language Identification
- Streaming ASR
- Offline ASR
- Forced Alignment
