Windows, Linux & Android · Free

Your AI.
Your machine.
Nobody else's.

SMARAN.AI is your personal AI assistant for PC and phone. Wake it by voice and just talk: it understands your screen, opens and controls your apps, plays your music, and builds websites, images and video for you. Run models locally, and your data never leaves your machine.

>

What it does

Everything in one window,
local models on your machine.

Assistants usually mean a browser tab and a monthly bill. This one is a program on your computer that happens to be very good at talking.

Talk to it like a person

A real-time voice session with interruption handling — start speaking and it stops to listen, the way a conversation actually works. Pick a voice and a persona; it answers in the language you set. The voice that runs on your machine speaks English; other languages are spoken by an online voice, so those need a connection.

It can see

Point the camera at a thing, or drop in a screenshot. Vision models read what is in front of them.

Your own documents

Hand it PDFs, notes, spreadsheets. It answers from those, and says so when the answer is not in them rather than inventing one.

It browses for you

Give it a task and watch a real browser window work through it - searching, reading, clicking - with every step shown. It never types passwords and stops before paying or sending anything.

"Hey SMARAN" - from any app

Say "Hey SMARAN", "Hey Amarya", "Hey Myra" or "Hey Jarvis" and it answers "How may I help you?" - play, pause, skip an ad, open an app, look at your screen, draw in Paint. On the phone all four names are heard offline; on the PC "Hey SMARAN" and "Hey Jarvis" are. Your music keeps playing. It never pays, buys or restarts anything without asking you.

Your phone and your PC, paired

Scan a QR code once. After that the conversation follows you between devices, and either one can drive the other — speak to your phone, and your computer does the work.

Locked like a phone

Enable a startup PIN in Settings, with attempt limits and lockouts to protect access to your history.

Real hardware readings

CPU, memory, GPU and battery from the actual machine — including whether it is charging — not a decorative animation.

Sixty-three models, one hub + HuggingFace GGUF

Download local models and keep them forever, search and pull quantized models straight from the built-in HuggingFace GGUF Hub, or plug in a key for a cloud provider (Gemini, Groq, OpenRouter, Cerebras, Mistral, DeepSeek, OpenAI, Anthropic and more). Delete local Ollama models permanently or remove cloud providers directly from Settings → Model Matrix. The hub is honest about the difference between catalogued, downloaded and actually running — and when no model answers, it says so plainly instead of inventing a reply or sending your chat to an unchosen service.

ReasoningVisionCodeTextAudioVideoGGUF Hub

It uses your computer

Give it a goal - "open Notepad and write the list" - and it looks at the screen, clicks and types with your mouse and keyboard, and looks again, every step shown live with its screenshot. Tell it something while it works and it changes course; in Notepad and other plain text boxes it reads the text back to check the typing really landed. Stop ends it at once. While a window that deletes files or changes disks is open, it will not click or type at all. It never types a password or an OTP, and anything about money, buying or the power button comes back to you as a question.

Answers with sources & Sparkpages

With Web on, SMARAN runs multi-source Deep Research across the web, papers, news, and forums — reading the best pages, checking every numbered citation against its source, and building rich Sparkpages with follow-up questions to tap.

A coding agent, not a chat box

SMARAN Code maps your repository (repo_map), makes atomic multi-file edits (multi_edit), runs Python & tests, and opens local web apps in its built-in browser to read console errors and verify UI changes. Pair it with Drive (Pages, Slides, Sheets, Podcasts), Dots, Mail, Team, and Dispatch in one workspace.

Pictures, video & music - free, in the cloud

Made in the free cloud: pictures from zero-key FLUX on Hugging Face (free AI Horde as backup) or a free NVIDIA FLUX key; short video from Wan 2.2 or LTX-Video on Hugging Face's shared free GPUs (with 5/10/15s continuation from the last frame, voice-over and soundtrack); and music from Lyria or MusicGen (plus a local PC synthesizer). Nothing in it costs money; the free cloud has a daily allowance, and busy hours mean a queue.

Free cloudImagesVideoMusic
The assistant

She has a face,
and she looks back.

The assistant is not a blinking cursor. Open Speak and she is standing there, breathing, watching, answering out loud — rendered live on your own machine, not streamed from anywhere.

  • Idle, thinking, speaking — she changes with the conversation
  • Pick a voice and a persona to match how you like to be spoken to
  • Rendered locally, so she works with the network off
Get her running
In action

Six things it does
that a chat box cannot.

01 — The workspace

One window, every mode

Chat, voice, vision and your documents are tabs of the same thing, not four separate apps. Switch mid-conversation and it keeps the thread.

  • Live hardware readings while it works
  • Artifacts render beside the conversation
  • Conversations kept per collection and per device
02 — Speak

A call, not a walkie-talkie

Press once and talk. It hears you while it is still speaking, and stops when you cut in — no push-to-talk, no waiting for a beep.

  • Interruption handled mid-sentence
  • Choose a voice and a persona
  • Answers in the language you picked
03 — Model Hub

Sixty-three models, plainly labelled

Filter by what you need — reasoning, vision, code, audio, video. Download a local one and it is yours forever, or point it at a cloud provider with your own key.

  • Says what each one needs before you download
  • Honest about catalogued vs. actually running
  • Local models cost nothing to run, ever
04 — Pairing

Your phone becomes a remote

Scan the code once. After that, speak to your phone and your computer does the work — or leave the computer off and the phone carries on alone.

  • Conversations sync in both directions
  • Only a vetted set of remote actions is allowed
  • Unpair from either device at any time
05 — Your documents

It reads what you give it

Drop in PDFs, notes or spreadsheets. Answers come from those pages — and when the answer is not in them, it says so instead of inventing one.

  • Scoped per collection, not one giant pile
  • Answers only from the collection you point it at
  • Files stay on your drive; with a cloud model, only the passages used for an answer are sent to it
06 — Autonomous Agents, WhatsApp & Analytics

Learns, and works while you are away

Second Brain memory across sessions, an interactive Executive Dashboards & CRM pipeline saved locally, a live 2-Way WhatsApp Linked Device QR gateway plus Telegram & Slack bots, real-time Deep Analytics telemetry, and scheduled background Dots. Commands run inside configurable Sandbox filters with checkpoints and one-click undo.

  • 2-Way WhatsApp Linked Devices QR inside SMARAN.AI (@smaran, /task, Message Yourself), plus Telegram & Slack
  • Live Deep Analytics (SQLite history, tokens/sec throughput, latency & workspace inventory)
  • Executive CRM with live KPIs, 6-stage funnel, tasks & Excel (.xlsx) export
  • Configurable Sandbox (STRICT / PERMISSIVE / OFF) & Capabilities live self-test
Getting started

Three steps, then it is yours.

  1. 01

    Install it

    On Windows, get it from the Microsoft Store; on Android, Linux or the command line, download it below. Python and the app are bundled, so there is no environment to prepare. To run models locally you also need Ollama, which is a separate free install.

  2. 02

    Pick a model

    Open the Model Hub and download one. A small one runs on almost anything; a bigger one wants more memory. It tells you which is which.

  3. 03

    Start talking

    Type, or press Speak. If you want it on your phone too, scan the pairing code.

Download

Free to download.

Every feature included, no card. Local models you download are yours to keep and cost nothing to run.

Latest Release: v1.1.84 (Stable)

Windows

10 & 11 · 64-bit · Microsoft Store

Get it from Microsoft Store

Signed by Microsoft (passes SmartScreen & Smart App Control), includes Energy Core, Myra & Amarya companions, supports built-in “Run on startup”, and auto-updates via the Store. (Press Speak or shortcut to talk; standalone installer includes offline wake words.)

  • The smallest models in the hub ask for 4 GB of memory
  • Built-in 2-way WhatsApp QR, Deep Analytics & Sandbox controls
  • A GPU makes it faster; it is not required
  • Webcam & mic optional for vision and live voice
  • Ollama runs local models; the app can install it for you

Android

7.0 and newer · v1.1.84 · 91.8 MB

Download APK
  • Answers on the phone itself — no computer or account needed
  • Works right away, or bring a free Gemini, Groq or OpenRouter key
  • Chats stay on the device, and read aloud
  • Pair a PC for documents, local models and control

After it downloads, open it from Downloads and allow installing.

Linux

64-bit · Ubuntu, Kali, Debian, Fedora, Arch, RHEL · Snap Store · v1.1.84

Get it from Snap Store

Install from the official Ubuntu Snap Store with sudo snap install smaran-ai — auto-updates via Snap Store and includes the built-in smaran CLI.

  • Works natively on Ubuntu, Kali Linux, Debian, Fedora, Arch, Manjaro, RHEL, Rocky, Alma, openSUSE, Linux Mint & Pop!_OS (glibc 2.28+)
  • Includes All Agents, Office Exporter (.pptx/.xlsx/.docx), Kanban CRM, SMARAN Code & 2-Way WhatsApp/Telegram/Slack
  • Runs local Ollama / HuggingFace GGUF models offline or connects your own cloud API keys
  • Chats, memory, documents and keys stay strictly on your machine — zero telemetry, zero ads
  • Opens in a clean Chrome, Chromium, Brave or Edge window; or your default browser

Also as .deb · .rpm · .tar.gz

Command line

Windows & Linux · v1.1.84 · 8.8 MB · no Python needed

Download Windows CLI Download Linux CLI
  • smaran ask, chat, models, status
  • Talks to the desktop app, so the app must be running
  • Local discovery uses loopback and your device account
  • One file. Put it on your PATH and it works
Documentation

How to install SMARAN.AI.

Step by step, for the platform you are on. Pick one.

  1. Open it in the Microsoft Store

    SMARAN.AI for Windows comes from the Microsoft Store. Open its Store page, or search for SMARAN.AI in the Store app.

    Go to downloads
  2. Click Get

    The Store installs it like any other app. It is signed by Microsoft, so there is no SmartScreen or Smart App Control warning, and new versions arrive by themselves.

  3. Install Ollama, if you want local models

    Models run through Ollama, which installs separately. Without it the app still runs and can use cloud providers with your own key.

  4. Open it

    First launch detects your hardware and suggests models that fit. Nothing downloads until you choose one.

SMARAN.AI Codex — VS Code extension

Work beside your code.

A coding agent in your editor’s sidebar. Give it a task and it reads your project, changes files, runs commands, reads what they printed and keeps going until the work is done — every step on screen, under a mode you choose.

Install the extension

Free and MIT-licensed.

Feedback

Send a review or a bug report.

Rate it, tell us what is missing, or report something that went wrong.

How was it?
not rated

Nothing is collected beyond what you type here. Leave the email blank and it stays anonymous.

Privacy

The honest version.

Most assistants say "we care about your privacy" and then send every word you type to a server. Here is what this one actually does.

✓ Kept on your machine

  • Your chats, memory and documents — no online account holds them
  • Your files, and what is in them
  • Dots and what they report
  • API keys — each one is sent only to the provider it belongs to
  • No usage data: the app reports nothing about you or how you use it
  • With a local model and the English voice, chat, voice and calls all work with the network off

↗ Leaves only for what you switch on

  • Cloud models — your message, and the context it needs, go to the provider you chose.
  • Web search — the search goes out when Web is on.
  • Reading aloud — English is spoken on this machine. Other languages send the reply’s text to Microsoft’s voice; the Gemini natural voice, if you choose it, sends it to Google.
  • Calls with Gemini Live — your voice, and your screen or camera if you share one, go to Google. Choose This computer in Settings → Voice and they do not.
  • Cloud pictures and video, 2-way WhatsApp Gateway (local Linked Devices profile), Telegram/Slack bots and connected apps — only what you send through them.
  • Android dictation uses the phone’s own speech service, which may work online.
  • AI dictation styles, Command mode and meeting Enhance — the words (never the audio) go to the cloud model you chose. Meetings are recorded and transcribed on your computer, and the audio is deleted once it is written down.
Questions

Questions, answered.

Is it free?

Yes. Every feature is in the free download. There is no paid tier, no usage cap and no card anywhere in the app. If you connect a cloud provider with your own key, that provider bills you directly — nothing to do with this app.

What is SMARAN.AI?

A free, private AI assistant made by Shashwat Mishra for Windows, Android and Linux. It chats, talks hands-free in English, Hindi and Hinglish, answers from the web with sources and from your own documents, and runs local AI models offline or cloud models with your own keys.

Where do I get it for Windows?

From the Microsoft Store. The Store copy is signed by Microsoft, so SmartScreen and Smart App Control let it run, and it updates itself. A downloadable .exe is no longer offered on this site.

Do I need internet?

To download the app and models, yes. After that, local models, speech recognition and the English voice work offline. Cloud models, web search, Gemini Live and the voices for Hindi and other languages need a connection. On Android, dictation uses the phone’s own speech service, which may need one too.

Does it speak Hindi and other languages?

Yes. It answers in the language you use — Hindi, Hinglish, Gujarati, Marathi, Tamil, Telugu, Bengali, Kannada, Malayalam and more — and reads each reply in a natural voice for that language. For calls you can also switch on Gemini Live (Settings → Voice & Speech) with your own Gemini key: Google’s real-time voice, which you can interrupt mid-sentence.

What are Dots?

Helpers that keep working for you: give one a goal — “tell me the important AI news every morning” — or pick a specialist (Researcher, Monitor, Writer or Developer), choose which tools it may use, and let it run on a schedule, from a webhook, on a matching incoming email, or when you message it from WhatsApp, Telegram or Slack. Each Dot has its own browser where you can sign in once and press Take over whenever it needs you. It asks before doing anything that changes something, and you can pause or delete it any time. With Keep running when the window is closed on, Dots carry on from the tray.

Can it take meeting notes?

Yes, without a bot joining your call. Open a page, press Record meeting, and type your own notes while SMARAN listens to your microphone and your computer’s sound — Zoom, Meet, Teams, anything. Press Stop and the transcript is written on your computer, marked You and Them. Enhance notes then turns your notes and the transcript into a summary, decisions and action items (general, 1:1, stand-up, sales call, interview or lecture). Ask a question in the Pages search box to get an answer from all your pages and meetings, with links to the ones it used. Please tell the others you are recording.

What hardware do I need?

It depends on the model, and the hub is specific about it. Across the 63 models in the catalogue the stated memory requirement runs from 4 GB to 512 GB, and 18 of them ask for 8 GB or less. Every entry shows its requirement before you download it. With a cloud key, an ordinary laptop is enough.

Can the phone app work when my PC is off?

Yes. The phone app is a working assistant on its own. Pairing adds shared history and lets each device drive the other, but it is not required.

Where is my data stored?

Chats, memory and documents are stored in your own user folder, and no online account holds them. The app sends no usage data. Something leaves your machine only for a feature you turn on — a cloud model, web search, Gemini Live, or a voice for a language other than English. Deleting the app’s data folder deletes everything.

How do I update?

Open Settings → Software Updates: it checks for a new version and gives you the download. Install it over the old one — your chats, models and settings stay.

Who built it

Built by one developer.

Stop renting your assistant.

Install it once. It is yours after that.