AI assistant for safe, reliable productivity with code, collaboration, and integrations.
Whisperstream
About Whisperstream
Whisperstream is a Windows application that provides private, on-device dictation without sending audio to the cloud. It transcribes speech in real time using a local NVIDIA Parakeet model, typing the output directly into any active application. The tool supports push-to-talk, global keyboard shortcuts, and hands-free toggling, with punctuation handled automatically. Users can import audio files for offline transcription and maintain a local transcript history that is encrypted on the device. The software includes features like a custom dictionary for names and jargon, per-app AI prompts, and a high-sensitivity Lockdown Mode to disable cloud enhancements and protect local data. It operates entirely offline after setup, with core transcription running on the CPU without requiring a GPU. The application is compatible with 39 languages and works across standard Windows applications like Outlook, VS Code, and browsers without requiring add-ins.
Key features
- On-device speech-to-text transcription
- Push-to-talk and global keyboard shortcuts
- Custom dictionary for names and jargon
- Per-app AI prompt configuration
- Local transcript history with encrypted storage
- High-sensitivity Lockdown Mode
- Audio file import for offline transcription
- 39 supported languages
Use cases
- Private dictation for long-form writing in applications like Gmail or Word
- Code documentation and comments in IDEs like VS Code
- Meeting notes and transcription from recorded audio files
Pros
- Local transcription with no cloud audio upload
- Supports 39 languages
- No subscription required
- Works offline after initial setup
- Encrypted local transcript history
Cons
- Windows-only platform
- One-time purchase model with no free tier beyond trial
- Limited to CPU-based transcription
Frequently asked questions about Whisperstream
What is Whisperstream?
Whisperstream is a Windows application that provides private, on-device dictation by transcribing speech in real time using a local NVIDIA Parakeet model. It types the output directly into any active application without sending audio to the cloud.
Who is Whisperstream suitable for?
Whisperstream is suitable for Windows users who prioritize privacy and want to dictate text directly into applications like Outlook, VS Code, or browsers without cloud processing. It benefits professionals, writers, and developers who need hands-free typing.
Does Whisperstream require an internet connection?
Core speech recognition and text delivery can run entirely offline after setup. Internet access is only required for license validation, updates, model downloads, and optional cloud-based enhancements you may configure.
How does Whisperstream handle pricing?
Whisperstream operates on a one-time purchase model with no subscription. A single license costs $29 and activates on one device, while a $49 license covers up to two devices. A free trial is available with no account or credit card required.
What languages does Whisperstream support?
Whisperstream supports 39 languages, including English, Spanish, French, German, Chinese, Japanese, Arabic, Hindi, and more. All transcription occurs locally without cloud dependency.
Can I use Whisperstream with multiple PCs?
Yes. You can choose a license size at checkout: a $29 license activates on one device, and a $49 license activates on up to two devices, such as a laptop and desktop.
Whisperstream Website Engagement
Last Update: 9 days ago
Monthly Traffic
Traffic Sources
Traffic Share By Country
- United States68.3%
- India29.5%
- Pakistan2.2%