↳ ZenonAI — initialising Samantha...
Six capabilities. One voice command. Zero cloud dependency.
Powered by Mistral, LLaMA 3, Falcon, Phi-3. Zero subscriptions — no data leaves your machine. Swap models with one line in .env.
Speak naturally. Whisper transcribes locally at near-human accuracy. Samantha listens, understands, and acts — no cloud relay.
Browse URLs, Google search, compose Gmail — driven by Playwright Chromium. Samantha sees your browser exactly like you do.
Open Spotify, send hotkeys, take screenshots. Full OS integration on macOS, Linux, Windows — no separate agent process needed.
Instant knowledge from Wikipedia and Google. Structured, summarised answers delivered conversationally in real time.
SQLite conversation history. Samantha remembers context across every session — building richer interactions over time.
“I've been using Samantha for months and have never seen technology like this. It feels like talking to a real assistant. Genuinely mind-bending.”
“Finally an AI agent that respects my privacy. Everything runs locally — no subscriptions, no data harvesting. Pure intelligence on my machine.”
“I replaced three paid tools with Samantha. Browser automation, email, Wikipedia — one voice command does it all. Absolutely incredible.”
“The memory feature is what hooked me. It remembered context from three days ago and used it perfectly. This is what AI was always supposed to be.”
“I've been using Samantha for months and have never seen technology like this. It feels like talking to a real assistant. Genuinely mind-bending.”
“Finally an AI agent that respects my privacy. Everything runs locally — no subscriptions, no data harvesting. Pure intelligence on my machine.”
“I replaced three paid tools with Samantha. Browser automation, email, Wikipedia — one voice command does it all. Absolutely incredible.”
“The memory feature is what hooked me. It remembered context from three days ago and used it perfectly. This is what AI was always supposed to be.”
“I've been using Samantha for months and have never seen technology like this. It feels like talking to a real assistant. Genuinely mind-bending.”
“Finally an AI agent that respects my privacy. Everything runs locally — no subscriptions, no data harvesting. Pure intelligence on my machine.”
“I replaced three paid tools with Samantha. Browser automation, email, Wikipedia — one voice command does it all. Absolutely incredible.”
“The memory feature is what hooked me. It remembered context from three days ago and used it perfectly. This is what AI was always supposed to be.”
“Running LLaMA 3 locally through Samantha is wild. Zero latency vs cloud APIs. I switched from GPT-4 and haven't looked back.”
“The open-source community needed exactly this. Full async, Playwright, voice I/O — it just works. The agent I always wanted to build.”
“Set it up with Mistral in under 10 minutes. Said 'open Spotify' and it just did. Still get goosebumps thinking about that first moment.”
“Voice recognition running completely offline at this accuracy is unreal. This is the future of desktop AI and it is already here today.”
“Running LLaMA 3 locally through Samantha is wild. Zero latency vs cloud APIs. I switched from GPT-4 and haven't looked back.”
“The open-source community needed exactly this. Full async, Playwright, voice I/O — it just works. The agent I always wanted to build.”
“Set it up with Mistral in under 10 minutes. Said 'open Spotify' and it just did. Still get goosebumps thinking about that first moment.”
“Voice recognition running completely offline at this accuracy is unreal. This is the future of desktop AI and it is already here today.”
“Running LLaMA 3 locally through Samantha is wild. Zero latency vs cloud APIs. I switched from GPT-4 and haven't looked back.”
“The open-source community needed exactly this. Full async, Playwright, voice I/O — it just works. The agent I always wanted to build.”
“Set it up with Mistral in under 10 minutes. Said 'open Spotify' and it just did. Still get goosebumps thinking about that first moment.”
“Voice recognition running completely offline at this accuracy is unreal. This is the future of desktop AI and it is already here today.”
Switch models with one line in .env — no code changes.
No Docker. No signup. No cloud. Just Samantha on your machine.
curl -fsSL https://ollama.com/install.sh | shLocal LLM server — runs on your machine.
ollama pull mistralOr: llama3 falcon:7b phi3 gemma2
curl -fsSL https://zenon.ai/install.sh | shInstalls Python deps, Playwright, and configures Samantha.
cd ~/samantha && python main.py --textRemove --text for full voice mode.
Or just run the installer
curl -fsSL https://zenon.ai/install.sh | shAutomatically installs Python dependencies, Playwright, and configures Samantha.
View install.sh source