Whisper is an open-source Android app and input method editor (IME) that brings OpenAI's Whisper speech recognition model to mobile devices for completely offline use. It supports both standalone transcription/translation and system-wide voice input via the keyboard. Models are downloaded once from Hugging Face; all processing happens locally. The app works with any compatible keyboard such as HeliBoard and includes features like editable results and language support.
In the AI space, Whisper takes a focused approach. It focuses on performing accurate speech-to-text transcription and translation completely offline on Android devices. Whisper is an open-source project aimed at android users needing offline dictation. Whisper costs nothing to use. It runs on the web and Android, and it can be self-hosted.
It is developed by woheller69, and it first shipped in 2024. The project is developed in the open on GitHub with 610 stars and 1 commit in the last 90 days. Among its 5 catalogued features are Offline Speech Recognition, Voice-to-Text IME, and Language Translation. The interface is available in 26 languages, including Arabic, Tibetan, and Catalan.
What PulseGate has recorded for this listing
Closest matches by what these projects do