The Subvocal Company, a Y Combinator F26 startup founded by Georgia Tech students Aarush and Arnav, has built prototype smart glasses that use radio-frequency sensing to decode silent or mouthed speech into text in real time. The founders say the device works even without visible lip movement, after earlier attempts using EMG facial electrodes proved unreliable due to sweat, interference and placement issues.
news.ycombinator.com
· 2026-11-02
Google has released an experimental app called Google AI Edge Foresight that transcribes meetings and audio files entirely on-device, using its EmbeddingGemma 2 model. The free app turns shorthand notes into polished summaries and includes a built-in assistant that can answer questions about meetings or connected local files, with Google saying data never leaves the user's computer. It currently only works on Macs with Apple Silicon.
theverge.com
· 2026-10-08
Fireflies.ai has launched a new feature called Fireflies Talk, letting users dictate text into any desktop app on Mac or Windows using a keyboard shortcut. The feature is included in existing subscription plans rather than sold separately, and the company says dictation transcripts are stored locally with no recordings kept.
techcrunch.com
· 2026-09-29
Wispr Advanced Interfaces Lab has released Canto, a new speech recognition model built specifically for real-time dictation in noisy, everyday conditions rather than clean studio audio. In tests using 10 hours of real Wispr Flow dictations from over 2,300 speakers, Canto posted the lowest word error rate compared with models from Google, OpenAI, AssemblyAI, and Deepgram.
wisprflow.ai
· 2026-09-17
London startup Relay, founded by former Nothing staffers Cookie Xu and Raymond Zhu, has released voice dictation software for macOS ahead of a companion hardware device called the Relay Q, due in early 2027. The founders say existing software tools like Wispr Flow struggled with inconsistent microphone pickup in office settings, prompting them to build purpose-made hardware for accurate speech capture.
wired.com
· 2026-08-27