Local / on-device and productivity AI tools

Local and on-device AI productivity tools for voice input and automation. - VoiceFlow: an open-source local speech-to-text tool using Whisper models. - VoiceFlow supports 16+ Whisper variants and 99 languages, local history search. - Voquill: cross-platform voice input tool with local Whisper and AI text-cleaning. - Voquill includes personal dictionary and optional GPU acceleration. - ClickClickClick: uses LLM vision to automate Android and macOS UI via natural language. - ClickClickClick maps high-level commands to device interactions using ADB and model vision. - Tools support local models via Ollama and major LLM providers for inference. - These open-source projects target users concerned with privacy and low deployment friction.
Key insights
VoiceFlow offers fully local Whisper-based speech recognition
VoiceFlow is an open-source desktop tool that runs entirely locally and supports multiple Whisper model sizes, 99-language recognition, and local searchable history. It pastes recognized text at the cursor and provides recording modes for quick input or long-form dictation, targeting users concerned about cloud privacy.
Voquill enhances cross-platform voice input with on-device cleaning
Voquill provides speech-to-text on macOS, Windows, and Linux with optional GPU acceleration and local Whisper models; it includes AI-powered text cleanup to remove filler words and a personal dictionary to reduce recognition errors for specialized terminology.
ClickClickClick enables natural-language device automation via LLM vision
ClickClickClick leverages large models' visual capabilities to automate Android and macOS interactions using natural language commands, translating instructions into ADB-driven actions or UI manipulations, and supports both cloud and local models such as Ollama.