Download Bodega One Code

One PowerShell command. No SmartScreen warning.

Open PowerShell (Win + X → Terminal), paste this, hit Enter. Downloads to %LOCALAPPDATA%\BodegaOne and launches. No admin rights, no installer dialog, no warning.

PS>irm https://bodegaone.ai/install.ps1 | iex

The script uses Microsoft's documented Unblock-File cmdlet to strip the Mark-of-the-Web that triggers SmartScreen. Read it first if you want: install.ps1 source.

Prefer a traditional installer?

Windows 10+ · 64-bit · ~180 MB

On a locked-down or corporate machine? Grab the portable zip. No installer, runs from anywhere.

macOS DMG

Two builds. Pick by your Mac's chip. Signed and notarized with our Apple Developer ID. Drag to Applications, open. Standard macOS install, no Gatekeeper warnings.

Apple Silicon = M1, M2, M3, M4 chips (most Macs since late 2020). Intel = pre-2020 Macs.

Linux AppImage

Single-file portable. Works on most modern Linux distros (Ubuntu 22.04+, Debian 12+, Fedora 38+, RHEL/Rocky/AlmaLinux 9+, Arch, openSUSE, Mint, Pop!_OS). CentOS 7 and 8 are end-of-life and not supported. Two builds. Pick by your CPU architecture.

x64 = standard Intel/AMD desktops + servers. ARM64 = Raspberry Pi 5, Apple Silicon Linux VMs, AWS Graviton, Asahi Linux.

Then run

$chmod +x Bodega-One-*.AppImage && ./Bodega-One-*.AppImage

System requirements

The app itself is light. The real hardware question is which AI model you run, and that is your call: local models need VRAM, cloud models need an API key and nothing else.

Running models locally? Size your hardware to the model with the VRAM calculator, or skip the question entirely with a cloud API key.

Verify your download

Every installer is hashed by GitHub when we publish it. Compare the SHA-256 of the file you downloaded against the list below before you run it. The installer does the same check itself on Windows, but you should not have to take our word for it.

Latest release: v1.0.0-beta.31.8.1 · published July 2, 2026

Bodega-One-Setup-latest.exe

8a1932726cf30949513e73ca52486c053c927ce0e841357f94324f20ccb6f1ae

Bodega-One-latest-portable-x64.zip

026ae6313d6cf8411021f7520e29c1ec1a85e8098301fbdde24d7fc2d198af2d

Bodega-One-latest-arm64.dmg

125d0ed9e255c347e9c5ef0b75f671bb20d9602cf7f176995ffc91ec81a760d8

Bodega-One-latest-x64.dmg

80efb73545bcad60f55417bfae9c22537973c79e371fc073597e3916d4586b0f

Bodega-One-latest-arm64.AppImage

8d4be76a4f80867ffed6ca288565960773887b991dd79d845ab71a3fb8cea0f8

Bodega-One-latest-x86_64.AppImage

2fe19c26f251f3b2da784c5e942f15592981f2d3f169be9af9aec616fa4b9cd2

Bodega-One-latest-amd64.deb

9da99001a8d428be4faf812dd8422e0ce96056033cac829571142ec3172256db

Bodega-One-latest-arm64.deb

0403054154d5beed97c9d8e7e3f875b31cd321494c2bcfcc078e74db9c0c783e

Windows (PowerShell)

PS>Get-FileHash .\Bodega-One-Setup-latest.exe -Algorithm SHA256

macOS / Linux

$shasum -a 256 Bodega-One-latest-*

Your first 10 minutes

1. 1Install and launchUse the one-liner or the installer for your platform above.
2. 2Activate with your emailFree. It ties updates and license recovery to you. Not telemetry; the app runs fully offline after activation.
3. 3Pick your model pathThe first-run wizard auto-detects running LLM servers. Have a GPU? Install a local model through managed Ollama or llama.cpp. Prefer cloud? Paste your own API key.
4. 4Open a project and put the agent to workStart with something small and reviewable. Every change lands as a diff with a QEL verification score.
5. 5Flip on air-gap if you need itZero network egress, enforced in the app, with an Air-Gap Active indicator in the top bar.

The full walkthrough lives in the Getting Started docs.

Looking for something else?

Older versions, checksums, release notes, and the full asset list live on GitHub.