Is internet required?
No. Model inference and data storage are both performed locally.
Assign dedicated recognition models to different speakers with automatic switching for higher accuracy.
Manage history records, speaker voiceprint library, and recognition results with easy access.
Custom dictionary auto-correction, batch delete and replace, AI correction
Built on multiple efficient ASR models, transcribe 30min audio in just 1min on CPU *
One-click install, one-click model download, graphical interface
All data processed locally, no internet required, sensitive info never leaks.
嗯，那么今天我们就简单的进行一下新生招聘的讨论吧...
嗯，地点的话我们现在可以有三个选择...
操场的话，这段时间太热了，我怕人流量有点少
确实，那考虑室内体育馆怎么样？
Support one-click launch, and intuitive model management, making complex large models simple and easy to use
Pre-configured for one-click translation, summarization, and correction
Tailor AI intelligence to your unique workflow with custom prompts
嗯，那么今天我们就简单的进行一下新生招聘的讨论吧...
Well, then we can discuss the recruitment of new graduates today...
嗯，地点的话我们现在可以有三个选择...
Well, the location options we have now are three choices...
我觉得我们可以把重点放在计算机学院那边...
I think we can focus on the Computer Science College...
No. Model inference and data storage are both performed locally.
Not required. The built-in speech-to-text models are optimized for CPU inference. Even a 10-year-old CPU can process 30 minutes of audio in about 3 minutes. When deploying large models with Ollama, a better GPU allows more advanced models.
Yes. Dual-channel simultaneous recognition is supported.
Mandarin (97%), Chinese dialects (90%), English (95%), Korean, Japanese, Italian (97%), Spanish (96%), Portuguese (95%), German (95%), French (95%), Russian (94%), Ukrainian (93%), Polish (93%), Dutch (93%), plus 25 other European languages.
Yes. It provides powerful editing features, including automatic custom-dictionary processing, click-to-play listen-and-edit mode, and batch modify/delete with automatic dictionary updates.
Supports major audio formats including MP3, WAV, FLAC, AAC, M4A, OGG, AIFF, ALAC, CAF, PCM, ADPCM, and WebM. Video or multi-channel audio can be converted with built-in tools before recognition.
Using the software may trigger an activation code requirement. However, if you have any reason to request free access, you can explain it in the feedback section and leave your email.