Automatic speech recognition for 52 languages and dialects, powered by Qwen/Qwen3-ASR-1.7B-hf. Upload or record audio to get a transcription with timestamps.