wan-video/wan-2.2-t2v-fast

A very fast and cheap PrunaAI optimized version of Wan 2.2 A14B text-to-video

Capabilities

Seed

Cost

Community model (estimated from hardware time)

Input Parameters

Name	Type	Description	Default	Constraints
`prompt`*	string	Prompt for video generation	`—`	—
`aspect_ratio`	string	Aspect ratio of video. 16:9 corresponds to 832x480px, and 9:16 is 480x832px	`"16:9"`	16:99:16
`disable_safety_checker`	boolean	Disable safety checker for generated video.	`false`	—
`frames_per_second`	integer	Frames per second. Note that the pricing of this model is based on the video duration at 16 fps	`16`	min: 5, max: 30
`go_fast`	boolean	Go fast	`true`	—
`interpolate_output`	boolean	Interpolate the generated video to 30 FPS using ffmpeg	`true`	—
`lora_scale_transformer`	number	Determines how strongly the transformer LoRA should be applied.	`1`	—
`lora_scale_transformer_2`	number	Determines how strongly the transformer_2 LoRA should be applied.	`1`	—
`lora_weights_transformer`	string	Load LoRA weights for transformer. Supports arbitrary .safetensors URLs from the Internet (for example, 'https://huggingface.co/Viktor1717/scandinavian-interior-style1/resolve/main/my_first_flux_lora_v1.safetensors')	`—`	—
`lora_weights_transformer_2`	string	Load LoRA weights for transformer_2. Supports arbitrary .safetensors URLs from the Internet. Can be different from transformer LoRA.	`—`	—
`num_frames`	integer	Number of video frames. 81 frames give the best results	`81`	min: 81, max: 121
`optimize_prompt`	boolean	Translate prompt to Chinese before generation	`false`	—
`resolution`	string	Resolution of video. 16:9 corresponds to 832x480px, and 9:16 is 480x832px	`"480p"`	480p720p
`sample_shift`	number	Sample shift factor	`12`	min: 1, max: 20
`seed`	integer	Random seed. Leave blank for random	`—`	—

promptrequiredstring

Prompt for video generation

aspect_ratiostring

Aspect ratio of video. 16:9 corresponds to 832x480px, and 9:16 is 480x832px

Default: "16:9"

16:99:16

disable_safety_checkerboolean

Disable safety checker for generated video.

Default: false

frames_per_secondinteger

Frames per second. Note that the pricing of this model is based on the video duration at 16 fps

Default: 16min: 5, max: 30

go_fastboolean

Go fast

Default: true

interpolate_outputboolean

Interpolate the generated video to 30 FPS using ffmpeg

Default: true

lora_scale_transformernumber

Determines how strongly the transformer LoRA should be applied.

Default: 1

lora_scale_transformer_2number

Determines how strongly the transformer_2 LoRA should be applied.

Default: 1

lora_weights_transformerstring

Load LoRA weights for transformer. Supports arbitrary .safetensors URLs from the Internet (for example, 'https://huggingface.co/Viktor1717/scandinavian-interior-style1/resolve/main/my_first_flux_lora_v1.safetensors')

lora_weights_transformer_2string

Load LoRA weights for transformer_2. Supports arbitrary .safetensors URLs from the Internet. Can be different from transformer LoRA.

num_framesinteger

Number of video frames. 81 frames give the best results

Default: 81min: 81, max: 121

optimize_promptboolean

Translate prompt to Chinese before generation

Default: false

resolutionstring

Resolution of video. 16:9 corresponds to 832x480px, and 9:16 is 480x832px

Default: "480p"

480p720p

sample_shiftnumber

Sample shift factor

Default: 12min: 1, max: 20

seedinteger

Random seed. Leave blank for random

Version: c483b1f7b892Updated: 7/25/2026190.9K runs