Speechify’s Windows app uses local models for transcription and dictation

Speechify, a leader in voice AI technology, has introduced a groundbreaking native Windows application that utilizes locally stored models for enhanced dictation capabilities across various applications. This new tool also offers the ability to read aloud articles, documents, and PDFs, tapping into Speechify's extensive library of voices. Competing against notable platforms like Wispr Flow, Willow, and Superwhisper, Speechify’s Windows application processes voice commands entirely on-device, specifically designed for Copilot+ PCs equipped with NPUs from major chip manufacturers such as AMD, Intel, and Qualcomm. Furthermore, it is compatible with other Windows 11 PCs that feature GPUs from these companies. The app operates three advanced models directly on the device: a neural text-to-speech engine, real-time voice activity detection, and transcription powered by Whisper technology. Users have the flexibility to switch to cloud-based models or adjust settings during use, providing a customizable experience. With a user base exceeding 50 million, Speechify employs the VITS Neural model, which can generate audio at seven different speed settings, making it easier for users to have content read aloud. Cliff Weitzman, the founder and CEO of Speechify, expressed enthusiasm about the launch, stating, "With over a billion Windows users globally, this release ensures that reading and writing are accessible regardless of the device or user preference. We are particularly excited about the enterprise potential, as many professionals have requested Speechify for their PCs." In addition to this launch, Speechify recently introduced a meeting transcription feature akin to Granola, although it was initially limited to web-based meetings. Now that the company has expanded its applications across platforms, there are plans to integrate this feature into native apps, allowing for transcription of meetings across any application or browser. Historically, Speechify focused primarily on text-to-speech functionalities, such as articulating articles, emails, and creating podcasts from documents. Recently, however, the company has been making strides towards becoming a comprehensive voice application, adding features like dictation, meeting transcription, and a voice assistant to its offerings.