Qwen3-ASR

•
11
•

Deploys a production-ready, all-in-one Docker solution for Alibaba's Qwen3-ASR, providing state-of-the-art automatic speech recognition across 52 languages with a dark-theme UI, REST API, and MCP integration.

Qwen3-ASR offers a comprehensive and production-ready solution for automatic speech recognition (ASR), packaged as an all-in-one Docker deployment. It leverages Alibaba's state-of-the-art Qwen3-ASR models to support 52 languages and dialects, featuring automatic language detection and precise word/character-level timestamps. The tool includes a user-friendly dark-theme UI for easy interaction, a robust REST API with Swagger documentation for programmatic access, and integration with MCP for AI agent workflows, all without any runtime downloads.

主な機能

015 GitHub stars
02Dark Theme Web UI with i18n (EN/CN/TW/JP)
0352 Languages & Dialect Support with Auto-Detection
04REST API with Swagger Docs & OpenAPI Schema
05Forced Alignment Timestamps (word/character level)
06Streaming Transcription via WebSocket

ユースケース

01Deploying a self-hosted, high-performance ASR service for production environments
02Developing multi-language speech-to-text applications with real-time transcription
03Integrating advanced ASR capabilities into AI agent workflows