What is Whisperstream?
Whisperstream is a dictation tool that works on both Windows PCs and Apple silicon Macs (currently in beta), converting spoken words into written text within various applications via a straightforward push-to-talk command. After installing the required speech-recognition models, users can dictate offline without needing an internet connection. The application supports several models, such as NVIDIA Parakeet, Qwen3 ASR, and Cohere Transcribe, and it also provides the flexibility to use the on-device Stream-1 Mini model or a preferred provider for enhanced processing. Whisperstream includes a variety of features like customizable vocabulary, text replacement capabilities, app-specific enhancement settings, and a searchable transcript history. Additionally, it can transcribe audio files and offers voice playback along with interactive responses when using Claude Code and Codex. The tool facilitates seamless dictation in various text fields without requiring extra add-ons, making it especially beneficial for users dictating in coding editors, emails, documents, and other applications. However, users should be aware that downloading models, performing software updates, and accessing certain online functionalities might require an internet connection for the best experience. Overall, Whisperstream aims to enhance productivity for anyone who prefers dictation in their daily tasks.
Pricing
Integrations
Screenshots and Video
Whisperstream desktop settings screen with transcription metrics, dictation language, voice model, and privacy controls.