uthana/text-to-motion-diffusion-v2

Generate 3D character animation data from a text prompt

Capabilities

Seed

Community model (estimated from hardware time)

Name	Type	Description	Default	Constraints
`character` *	string	Output character	`—`	Tar Ava Manny Quinn Y Bot
`fps` *	integer	Output frame rate	`—`	24 30 60
`output_ext` *	string	Output file type	`—`	FBX GLB
`prompt` *	string	Describe the desired motion	`—`	—
`cfg_scale`	number	Cfg Scale (higher values will attempt to follow the prompt more)	`1`	min: 1, max: 4
`foot_ik`	boolean	Enable Feet Inverse Kinematics	`true`	—
`internal_ik`	boolean	Use internal IK to improve result	`true`	—
`motion_length`	number	Desired motion length (seconds)	`5`	min: 0.2, max: 9
`no_mesh`	boolean	Should the output contain animation only	`false`	—
`seed`	integer	seed (0 indicates a random seed)	`0`	min: 0, max: 1000000000

character required string

Output character

Tar Ava Manny Quinn Y Bot

fps required integer

Output frame rate

24 30 60

output_ext required string

Output file type

FBX GLB

prompt required string

Describe the desired motion

cfg_scale number

Cfg Scale (higher values will attempt to follow the prompt more)

Default: 1 min: 1, max: 4

foot_ik boolean

Enable Feet Inverse Kinematics

Default: true

internal_ik boolean

Use internal IK to improve result

Default: true

motion_length number

Desired motion length (seconds)

Default: 5 min: 0.2, max: 9

no_mesh boolean

Should the output contain animation only

Default: false

seed integer

seed (0 indicates a random seed)

Default: 0 min: 0, max: 1000000000

Version: 0ad729c8d6e6 Updated: 6/26/2026 92 runs