Animations and scenarios

An emote is an animation played on your character: a wave, a dance, sitting on the ground. The game does not keep every animation in memory, so before you can play one you have to ask for it, wait for it to load, then play it. This lesson builds that exact sequence as a resource you can run, and shows the two ways to move a ped: a single animation clip and a built-in scenario.

You'll build: qu_emotes: a resource with /dance and /sit that plays a real animation and a stock scenario on your ped
Time: ~26 minutes
You need: A working resource you can restart and the RegisterCommand basics.
You'll learn: RequestAnimDict -> wait for it to stream -> TaskPlayAnim -> the flag numbers -> TaskStartScenarioInPlace -> ClearPedTasks to stop

BEFORE YOU START

Build it

Make the resource folder

The server has one folder for this lesson.

Inside your server's resources folder, create this folder:

code

resources/qu_emotes

Create the files

Every file named in the manifest exists.

Create this exact file layout. Emotes run on the player's machine, so this is a client-only resource.

code

resources/qu_emotes/
fxmanifest.lua
client.lua

Write fxmanifest.lua

FiveM knows which file to load.

Open fxmanifest.lua and paste this:

code

fx_version 'cerulean'
game 'gta5'

client_script 'client.lua'

Older tutorials add a lua54 'yes' line here. As of June 2025 that setting is deprecated and ignored: Lua 5.4 is now the only Lua runtime, so you leave it out.

Write the dance command

A reusable helper streams a dict and plays a clip.

Open client.lua and paste this. The playAnim helper is the whole pattern in one place: request, wait, play.

code

-- Request a dictionary, wait until it streams in, then play the clip on the player.
local function playAnim(dict, name, duration, flags)
RequestAnimDict(dict)
while not HasAnimDictLoaded(dict) do
    Wait(0)
end

local ped = PlayerPedId()
TaskPlayAnim(ped, dict, name, 8.0, 8.0, duration, flags, 0.0, false, false, false)

-- We have the data now, so let the engine free the memory when it can.
RemoveAnimDict(dict)
end

RegisterCommand('dance', function()
-- flags = 1 means the animation loops until something stops it.
playAnim('anim@amb@nightclub@dancers@crowddance_facedj@hi_intensity', 'hi_dance_facedj_09_v2_female', -1, 1)
print('[qu_emotes] dancing - run /stop to cancel')
end, false)

Add the sit scenario and a stop command

The ped can sit using a stock scenario and stop on demand.

Add these to the bottom of client.lua. A scenario is a packaged behaviour the game already ships, so it needs no dict and no flags.

code

RegisterCommand('sit', function()
local ped = PlayerPedId()
-- Scenarios are built-in behaviours; no dict to request, no flags to pass.
TaskStartScenarioInPlace(ped, 'PROP_HUMAN_SEAT_CHAIR_MP_PLAYER', 0, true)
print('[qu_emotes] sitting - run /stop to cancel')
end, false)

RegisterCommand('stop', function()
local ped = PlayerPedId()
-- ClearPedTasks ends whatever animation or scenario is currently running.
ClearPedTasks(ped)
print('[qu_emotes] stopped')
end, false)

Start and test it

The expected proof appears in F8 and your character moves.

Open server.cfg and add this line:

code

ensure qu_emotes

Save, then run:

code

restart qu_emotes

Join the server, open F8, and run these one at a time, watching your character each time:

code

/dance
/sit
/stop

/dance makes your character dance in a loop. /sit drops them into a seated scenario. /stop clears whichever one is running and your character stands back up to idle. If the dance fires but your character never moves on the first try, that is the streaming gap the next section is built to explain.

In-game · F8 consoledance

> /dance

[qu_emotes] dancing - run /stop to cancel

Keep reading the full lesson

Still ahead in this lesson

How it works
If something went wrong
What you can do now
Try it yourself