Whisper-WebUI
Making local speech-to-text and translation easy.
Features
System Requirements
16GB RAM recommended. 24GB+ storage recommended.
macOS 15+: Supports both Intel and M-series chips.
Windows 10/11: Intel/AMD GPUs supported, NVIDIA GPU recommended.
Note: For NVIDIA GPUs, install a newer driver.Introduction
Whisper-WebUI is a web-based graphical user interface (GUI) developed for the renowned speech recognition model, OpenAI Whisper. Its core purpose is to enable everyday users, even those without programming experience, to easily perform transcription (speech-to-text) and translation tasks on audio and video files through an intuitive browser window. The project is built upon the powerful OpenAI Whisper model and uses the Gradio library to quickly create a user-friendly web interface. It also supports integration with optimized, faster versions like faster-whisper.
Core Features: What can it do for you?
For beginners, its features are both powerful and straightforward:
One-Click Transcription with Ease
- Multiple Sources: You can upload local audio/video files, directly paste a YouTube video link, or use your computer's microphone for real-time recording.
- Multiple Formats: The generated text can be saved in common subtitle or text formats like SRT, VTT, TXT, making it easy for subsequent editing or use in video production.
Efficient Translation, Breaking Language Barriers
- Speech-to-Text Translation: It can directly transcribe and translate foreign language speech (e.g., French, Japanese) into English text in one step.
- Subtitle Text Translation: Supports uploading existing subtitle files and translating them into other languages using the integrated DeepL API or the NLLB model.
Smart Processing for More Accurate Results
- Integrated Voice Activity Detection (VAD) intelligently splits long audio into sentences, effectively avoiding the "hallucination" issues common in traditional tools during silent or noisy segments, resulting in cleaner transcripts.
- Supports Speaker Diarization, identifying and labeling speech segments from different people, which is particularly useful for meeting minutes or interview transcripts.
Project Characteristics: Why Choose It?
- Low Barrier to Entry: Requires just a few clicks, no need to understand the command line or code.
- Comprehensive Functionality: Integrates a complete workflow from input, transcription, and translation to output.
- Privacy-Friendly: Can be deployed locally on your own computer, meaning all audio data stays offline and isn't sent to third-party servers.
- Highly Customizable: As an open-source project, advanced users can choose between different Whisper model engines to balance speed and accuracy according to their needs.