In the rapidly evolving world of speech-to-text technology, Whispering is a new entrant with a clear commitment to openness and user control.
Developed under the Epicenter umbrella, an ecosystem of open-source, local-first apps, the tool stands out for its free, open-source nature, and a design that keeps data firmly on users’ devices.
Whispering is positioning itself as a transparent and cost-effective alternative to subscription-based speech-to-text platforms.
The app allows users to record and transcribe audio using either local models (Speaches) or cloud providers, including OpenAI Whisper, Groq, and ElevenLabs. Importantly, the tool connects directly through the user’s own API keys or operates entirely offline, ensuring no third-party servers process the data.
This design reflects the project’s central principle: users own their data.
Local transcription is free, while cloud-based transcription starts at roughly USD 0.02 per audio hour, with the project framing typical outlay at USD 0–5 per month because users “pay providers directly.” For many users, this equates to a monthly spend of just a few dollars, compared to the USD 10–30 subscription fees common among transcription platforms.
With customizable keyboard shortcuts that enable push-to-talk dictation into whatever app the user already has open, Whispering can be easily embedded into daily activities, from dictating meeting notes to supporting coding workflows.
Whispering also includes optional AI processing steps, such as formatting or light cleanup, that can be chained into simple, repeatable text workflows.
The app is available as a native desktop download for macOS, Windows, and Linux, plus a browser-based version for quick trials. A setup tutorial on YouTube demonstrates how users can quickly integrate the tool into their processes.
Please note that Whispering is designed for quick transcriptions, not long recordings. For extended recording sessions, the use of a dedicated recording app is recommended.
