File size: 2,002 Bytes
2695555
 
 
 
 
 
 
 
 
 
 
 
 
 
7b82088
 
2695555
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7b82088
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
---
license: mit
tags:
  - assistant
  - local-ai
  - desktop
  - tool-use
language:
  - en
  - pt
---

# openTARS

<img src="https://huggingface.co/spaces/OpenTARS/README/resolve/main/banner.png" alt="openTARS banner" width="100%">

A local AI assistant for Linux, Windows and macOS. Ask in your own words and it opens programs, clicks buttons by name, types, searches the web, looks at the screen and runs commands. With a local model, processing stays on your machine. No subscription.

This repository is a **mirror** of the app for people who find openTARS here. The source of truth, issues and releases are on GitHub: https://github.com/project-opentars/OpenTARS. Website: https://project-opentars.github.io/OpenTARS/

## What is here

- `opentars_all.deb`: the Ubuntu/Debian package (version 5.2 "Ranger", revision 2026110101).
- `app/`: the application source and the installers (`app/empacotamento/`).
- `LICENSE`: MIT.

## Install

Windows 10 or 11, in PowerShell:

```powershell
irm https://raw.githubusercontent.com/project-opentars/OpenTARS/main/app/empacotamento/windows/instalar.ps1 | iex
```

macOS 12 or newer, in Terminal:

```bash
curl -fsSL https://raw.githubusercontent.com/project-opentars/OpenTARS/main/app/empacotamento/macos/instalar.sh | bash
```

Ubuntu, Debian and derivatives:

```bash
wget -O /tmp/opentars.deb https://github.com/project-opentars/OpenTARS/raw/main/opentars_all.deb && sudo apt install -y --reinstall /tmp/opentars.deb
```

The Windows and macOS installers are still early and have only been tested in automated runs and on one Windows laptop. Read the script before running it: the install commands above download and run code from GitHub.

## Models

- [Gargantua 2](https://huggingface.co/OpenTARS/Gargantua-2): the main model (2B, with vision).
- [Gargantua 2 Lite](https://huggingface.co/OpenTARS/Gargantua-2-Lite): the 0.8B model that runs on the CPU.
- [Murph](https://huggingface.co/OpenTARS/Murph): the tiny request classifier built into the app.