Install · Your PC
Get Companion & XenLook 4b-Ko
Install like a game: Companion plus the 4B model on your machine. Daily chat runs on localhost; saves stay in your folder only.
Web chat is for trials. For voice, local models, privacy, and near-offline use, install Companion — Ollama first, then the setup wizard in on chat.
Download CompanionGuided path needs no terminal — folder pick, model check, and daemon verify in your browser.
- 8 GB+ RAM recommended (16 GB for smoother local 4B)
- ~3 GB free disk for XenLook 4b-Ko via Ollama
- Ollama installed and running (Windows · macOS · Linux)
One-click install — 3 steps
Like installing a game: Ollama → model → on wizard. Terminal only on Linux advanced path.
Ollama runs the AI model on your PC — like a game engine. Download, install, and leave it running in the background.
Signed .msi/.dmg pending certificates — use Ollama + setup wizard for now.
First time? Follow these 5 steps
About 15 minutes · no coding required · Windows, Mac, or Linux
- 1
Install Ollama (free app)
Ollama runs the AI model on your PC — like a game engine. Download, install, and leave it running in the background.
- 2
Open the setup guide in on chat
Log in at on.xenlook.com. We walk you through folder, model, and connection checks in plain language.
Download Companion - 3
Pick where chats are saved
Choose a folder on your PC (e.g. Documents/xenlook). Chats stay there — not on our server.
- 4
Download the Korean 4B model
One copy-paste command or a button in the guide. Needs ~3 GB disk space.
- 5
Press “Test connection”
When both tests turn green, daily chat runs on your PC. You can still use cloud mode anytime.
Start on your operating system
Signed store installers coming soon. For now Ollama + on guide is the easiest path.
Linux
① Ollama or auto-install script ② on setup guide ③ Connection test
What local chat feels like
Companion (optional) — install like a game on your PC. Saves stay in your folder; everyday chat runs on your GPU locally.
Stuck? Common questions
What is Ollama?
A small free app that runs XenLook’s model on your computer. Install it once, keep it open in the background.
Do I need the terminal?
No for the main path. The on chat guide uses buttons. Terminal is only under “Advanced”.
Test failed — what now?
Make sure Ollama is running, wait 30 seconds, try again. Still red? Use Advanced install or ask in on chat.
Sensitive data stays off servers
Companion local vault turn plaintext is not stored in XenLook server DB (⑧·⑨). Cloud chat applies block-and-delete for sensitive patterns such as national ID or bank account numbers (AI Policy §3). PC security and backups are your responsibility.
Faster everyday chat
No cloud round-trip — 4B inference on localhost. Short replies without internet wait or shared server queue.
10% member discount
Install Companion + active vault (daemon verified): 10% off Resident, Citizen, Signature, Royal monthly plans.
Default is web on (cloud). Companion is optional; local mode never turns on without consent. 10% paid-member discount applies after vault + daemon heartbeat verification. Cloud used only on opt-in reasoning turns.
Full technology & privacy details →Regional local models
Korea ships with XenLook 4b-Ko today. Americas, EU, and Asia (ex-KR) 4B lanes target CDN hosting in 2026 Q4.
Selected: Japan, United States & others · Gemma 4 4B
Gemma lane: local Gemma model. EU Mistral (Vertex) is not connected — quality limits are disclosed honestly.
manifest: https://xenlook.com/download/companion/models/americas/manifest.json
| Region | Model | Size | Status |
|---|---|---|---|
| kr | XenLook 4b-Ko xenlook-4b-ko | 4B | Ready |
| asia | Qwen 3.5 4B qwen3.5:4b | 4B | Ready |
| americas | Gemma 4 4B gemma4:e4b | 4B | Ready |