CJ Pais
CJ Pais is an independent software artist and developer focused on technology that gets out of the way, helps people live more presently, and supports authentic connection. His work emphasizes private, local-first and open-source software, particularly accessible speech-to-text tools.
At a Glance
- Individuals seeking free private dictation
- Accessibility and assistive-technology users
- Open-source developers
- Desktop application developers embedding local ASR
- +2 more
AI Tools by CJ Pais
(1)
Handy
Open Source Offline Dictation App
Discussions
No discussions yet
Be the first to start a discussion about CJ Pais
Latest News
Handy repository updates added language detection support and fixed Intel macOS Metal builds.
Handy v0.9.8 released with bug fixes, a transcribe.cpp 0.2.4 update, Chinese script selection and shortcut/recording reliability improvements.
Handy v0.9.7 released with audio-tail fixes, experimental voice-activity detection, auto push-to-talk, debug UI, new translations and transcribe.cpp 0.2.3.
Handy v0.9.6 released with Linux paste fixes, overlay improvements, custom-word fixes, and transcribe.cpp 0.2.0.
Products & Services
Free, open-source and extensible desktop speech-to-text application. Users press a configurable shortcut, speak, and Handy transcribes locally and pastes the result into the active text field.
A ggml-based local transcription/inference library supporting 16 ASR families and more than 60 models, GPU acceleration, streaming and batch transcription, and maintained Python, JavaScript/TypeScript, Rust and Objective-C/Swift bindings.
A 1985 Apple M0100 mouse converted into a wireless speech-to-text button.
A 3D-printed foot pedal for speech-to-text control.
Market Position
CJ's work is positioned as a privacy-first, free and open-source alternative to cloud dictation and paid tools such as MacWhisper, Superwhisper, VoiceInk and Wispr Flow. Handy differentiates through local inference, cross-platform availability, extensibility and a deliberately simple 'speak into any text field' workflow; transcribe.cpp targets developers who need validated, accelerated and embeddable ASR rather than a hosted API.
Founding Story
CJ Pais is an individual/solo developer rather than a company. He left Qualcomm to pursue technology full-time and has focused on independent and open-source projects. Handy began after he broke a finger and needed a more flexible way to control his computer; he built it as a simple, extensible, local speech-to-text tool and released it openly.
Business Model
Revenue Model
Handy is free and open source. CJ supports the work through donations and GitHub Sponsors; transcribe.cpp also benefits from project support and in-kind credits from Mozilla AI, Modal and Blacksmith.
Pricing Tiers
Open-source desktop application with no paywall; donations and sponsorships are invited.
Target Markets
- Individuals seeking free private dictation
- Accessibility and assistive-technology users
- Open-source developers
- Desktop application developers embedding local ASR
- Researchers and engineers working with speech-recognition models
- Users of macOS, Windows and Linux
- Private offline dictation in browsers, editors, chat apps and forms
- Accessibility and alternative input for people who cannot comfortably type
- Hands-free writing, coding and communication
- Multilingual local speech recognition
- Embedding local ASR into desktop and mobile applications
- Developing and benchmarking accelerated speech-recognition applications
- Handy is used directly by open-source users and accessibility-focused users; the project site identifies sponsors Wordcab, Epicenter, Bolt AI and Canty Digital.