computer

tachyonlab

About this app

Your conversations hold valuable information — decisions, ideas, details you'll forget. Computer captures them and makes
them searchable with AI, entirely on your device. No cloud. No uploads. No one listening but you.

▎ RECORD & TRANSCRIBE LOCALLY
▎ Tap once to record. When you stop, Computer transcribes your audio using Whisper speech-to-text — right on your phone.
Choose from 9+ transcription models, from ultra-fast Tiny (40 MB) to studio-grade Large V3 (1 GB+). Your audio never
leaves your device.

▎ ASK YOUR MEMORY
▎ Type a question in the terminal and Computer searches your entire transcript library using AI-powered semantic search
(RAG). It doesn't just match keywords — it understands meaning. Ask "what did we decide about the budget?" and get an
answer pulled from the right conversation, with context.

▎ CHOOSE YOUR AI
▎ Pick the on-device language model that fits your phone. 9 models to choose from:
▎ • Gemma 4 E2B/E4B — Google's latest multimodal models
▎ • Gemma 3 1B — best balance of speed and quality
▎ • DeepSeek R1 — advanced chain-of-thought reasoning
▎ • Qwen 2.5 — strong multilingual support
▎ • SmolLM 135M — lightning fast on any device
▎ • Phi-4 Mini — top quality from Microsoft

▎ PRIVATE BY DESIGN
▎ • All AI runs on-device — no internet required after setup
▎ • No audio uploaded anywhere, ever
▎ • Biometric lock (fingerprint / face ID) protects your data
▎ • Local SQLite database — you own your data completely
▎ • No tracking, no analytics, no accounts

▎ BUILT FOR POWER USERS
▎ • Real-time streaming AI responses
▎ • Full transcript details with word count, language detection, copy-to-clipboard
▎ • Storage dashboard showing model and database usage
▎ • Download only the models you need, swap anytime

▎ Perfect for meetings, lectures, interviews, brainstorming sessions, or any conversation worth remembering. Computer
turns your phone into a private, searchable second brain

What’s new

Stability fixes