​Audio Transcription and Recording Tool
Version 1.0.0 - Freeware application​
To use this tool, please read the instructions below carefully and watch the tutorial video. If it improves your workflow, please consider supporting its development.
Main Features:
High-Accuracy Audio Transcription
-
Convert your .mp3 and .wav files into text effortlessly.
-
Powered by OpenAI Whisper, ensuring state-of-the-art transcription quality.
-
Supports 8 major languages, including Portuguese, English, Spanish, French, German, Italian, Russian, and Chinese.
"AudioT simplifies your workflow, whether you're a student, researcher, educator, journalist, or professional who needs accurate and fast audio-to-text conversion. It also integrates well with your preferred AI tools."
​
Alexandre Oliveira, Developer
Fast Processing with GPU Acceleration
-
Optional GPU support for lightning-fast transcriptions, saving your precious time.
Flexible Whisper Model Options*
-
Choose the model that fits your needs and resources:
Tiny: Super fast, minimal resource usage.
Base: Balanced speed and accuracy.
Small: Reliable accuracy for general transcriptions.
Medium: High accuracy for clear audio.
Large-v3: Maximum accuracy for complex tasks.
Built-In Audio Recorder
-
Record audio directly within AudioT with Start, Pause, and Stop controls.
-
Save recordings as .wav or .mp3 files, ready for transcription or sharing.
-
Experience visual feedback with the blinking record button for active recording.
Instant Transcription Access
-
Save transcriptions as .txt files with one click.
-
View texts in a resizable window with easy copy-to-clipboard functionality.
-
Quickly locate your files with the Folder button.
Multilingual & User-Friendly Interface
-
Switch between 8 interface languages seamlessly.
-
Supports light and dark modes for your comfort.
-
Stay updated with real-time progress bars.
-
Built with robust error handling for smooth, worry-free usage.
User interface: light mode.
Basic Requirements for Installation
-
Works on Windows devices.
-
The minimum required disk space is 2.64 GB for the default installation and 6.94 GB for installation with GPU acceleration.
Instructions:
1. Preferably, install and run as administrator: during installation, Windows may display a security warning because AudioT does not yet have a paid digital signature. This is normal for independent software and does not indicate any risk. You can safely proceed with the installation
​
​​​2. In this version, the “small.pt” and “medium.pt” multilingual model files are included. They are suitable for most cases and offer the best cost-benefit in terms of speed and accuracy. If you need a different model, you can download it below and manually place the *.pt file into the installation folder at C:/.../AudioT/_internal/whisper/models/. Alternatively, log in to Hugging Face at https://huggingface.co/login, search for Whisper models, and download the desired models from their details page.​
​
3. The small model performs efficiently in most cases with clear audio, providing faster transcription, while the medium model takes longer but delivers higher accuracy. If needed, you can download and test other models.
​
4. Saving audio as .mp3 produces significantly smaller files than .wav and allows for faster transcription.​​
​
5. Longer audio files can be time consuming during the transcription process. If the status bar pauses at about 80%, please wait, it is normal.
User interface: dark mode
How to use AudioT tool - Tutorial Vídeo (Soon...)
If AudioT makes your work more productive, please consider making a small donation. Your support helps keep it free for everyone and enables its continued development. Every contribution makes a difference, and it only takes a few seconds.
AudioT Installer (1.87 GB)
MediaFire
Mega
Download Whisper Models
MediaFire
Mega