Windows
x64
Package format: setup.exe
Installer package with one-click startup and Free 10-seat access.
Install
Deploy your first private AI collaboration pilot in minutes.
Use this hub to pick your operating system, follow first-run steps, validate one real workflow, and prepare for secure team rollout.
Before you install
- A machine in your internal network (desktop or server) with outbound access to download release assets.
- One pilot owner (admin) and at least one teammate account for first workflow validation.
- A model runtime plan (for example local Ollama setup) before multi-agent tests.
macOS
Universal
Package format: dmg
Includes Apple Silicon and Intel support with Free 10-seat access.
Linux
x86_64
Package format: tar.gz
Single binary package for server deployment with Free 10-seat access.
First-run verification checklist
- 1Install package for your operating system and launch 123AIChat.
- 2Open local URL, create admin account, and complete basic setup.
- 3Create one channel and invite one teammate or bot account.
- 4Send one message and trigger @bots to verify streaming response flow.
- 5Capture pilot notes: installation time, stability observations, and next rollout scope.
Troubleshooting map
App starts but no AI reply
Verify model readiness and routing order (global -> channel -> bot).
Channel has traffic but members miss messages
Check membership scope and channel assignment before testing broadcast.
Pilot team unsure if setup is production-ready
Run one repeatable scenario and record outputs in an internal checklist.
Frequently asked questions
How long does a first private pilot usually take?
Most teams complete installation and first workflow validation in about five minutes, then spend additional time on security and routing checks.
Can we start from a local desktop and later move to server deployment?
Yes. A desktop pilot is common for validation, then teams transition to internal server environments for broader rollout.
What should we verify before inviting more teams?
Confirm channel isolation behavior, model routing stability, and one auditable workflow outcome from your pilot channel.