Navigation
Recherche
|
Reactorcore Announces Free Offline Speech-to-Text Tool Using OpenAI Whisper
vendredi 11 juillet 2025, 05:54 , par KVR Audio
Reactorcore has announced a new free tool called Whisper Batch Transcriber, bringing professional-grade speech-to-text capabilities to Windows users without requiring internet connectivity or subscription fees. The tool leverages OpenAI's Whisper AI technology to automatically convert audio files into clean, formatted text documents. Unlike cloud-based transcription services, it runs entirely offline on the user's computer, ensuring privacy and unlimited usage without recurring costs. Key Features: 100% offline operation - no internet required after setup. Supports multiple audio formats: MP3, WAV, FLAC, OGG, M4A, AAC. Two AI models included: Small_en (2GB VRAM, 96% accuracy) and Large_v3_Turbo (6GB VRAM, 99.8% accuracy). Batch processing capability for multiple files. Clean, formatted output with proper punctuation and capitalization. The tool is particularly useful for musicians, podcasters, content creators, and audio professionals who need to transcribe interviews, voice memos, rehearsal recordings, or other audio content. Processing time varies by hardware, typically taking 2-10 minutes per minute of audio. System Requirements: Windows 7 or newer. Python 3.10+ (automated installer included). Minimum 2GB GPU VRAM (6GB+ recommended for best accuracy). 2GB storage space. The developer notes that while the tool takes 5-15 minutes to initially load, this is normal behavior. For optimal performance, audio files should be converted to mono WAV format at 16kHz sample rate beforehand. Python source code is included in the download too. If you have Python installed, you can run the application slightly faster by double-clicking the.py file instead of the.exe. Whisper Batch Transcriber is available as a free download from the developer's itch.io page, with optional support through donation for users who find the tool valuable. Read More
https://www.kvraudio.com/news/reactorcore-announces-free-offline-speech-to-text-tool-using-openai-wh...
Voir aussi |
126 sources (21 en français)
Date Actuelle
sam. 12 juil. - 13:19 CEST
|