Skip to main content
This guide takes a clean FiveM server from registration to its first AI NPC.

​
1. Register the server

Open the AI NPC dashboard, choose Register server, and enter:
  • A name that identifies the server in your dashboard.
  • Its globally routable public IPv4 callback URL, for example http://PUBLIC_IPV4:30120.
Use the public IPv4 address players use to reach the FXServer. Hostnames, IPv6, private addresses, loopback addresses, and reserved example addresses are not accepted. The URL must use http, contain an explicit unprivileged port (1024–65535), and have no path, query, or fragment. The dashboard shows the new server license once. Copy it immediately. You can rotate a lost or exposed license later.

​
2. Download the resource

Download humalike.zip from the latest GitHub release. The archive contains a ready-to-run humalike directory. Extract it into your server resources:

​
3. Configure FXServer

Add the license immediately before the resource starts:
Use set, not setr, for humalike_license_key. A replicated convar would expose the license to connected clients.
Do not configure voice URLs or voice secrets. HumaLike provisions voice automatically from the server license.

​
4. Verify the connection

Start the server and run this in its console:
The resource should report installed runtime credentials and a selected player provider. Framework and inventory providers appear when their dependencies are running.

​
5. Create an NPC

In the dashboard:
  1. Open the registered server.
  2. Select Create NPC.
  3. Start with a Static NPC.
  4. Choose its name, personality, language, voice, model, and coordinates.
  5. Save and enable it.
The NPC appears after the next roster refresh. You can also restart only the humalike resource while setting up a new server.

​
6. Talk

Join the server, approach the NPC, hold your configured voice push-to-talk key, and speak. HumaLike selects the NPC you are addressing from proximity, view direction, following state, and vehicle context.
If the NPC appears, shows an active voice indicator, and answers, the core installation is complete.
If any step fails, see Operations and troubleshooting.

​
Next