wan-video/wan2.1-with-lora

Run Wan2.1 14b or 1.3b with a lora

Capabilities

Reference ImagesNegative PromptSeed

Cost

Community model (estimated from hardware time)

Input Parameters

Name	Type	Description	Default	Constraints
`prompt`*	string	Text prompt for video generation	`—`	—
`aspect_ratio`	string	The aspect ratio of the video. 16:9, 9:16, 1:1, etc.	`"16:9"`	16:99:161:1
`fast_mode`	string	Speed up generation with different levels of acceleration. Faster modes may degrade quality somewhat. The speedup is dependent on the content, so different videos may see different speedups.	`"Balanced"`	OffBalancedFast
`frames`	integer	The number of frames to generate (1 to 5 seconds)	`81`	1733496581
`image`	string(uri)	Image to use as a starting frame for image to video generation.	`—`	—
`lora_strength_clip`	number	Strength of the LORA applied to the CLIP model. 0.0 is no LORA.	`1`	—
`lora_strength_model`	number	Strength of the LORA applied to the model. 0.0 is no LORA.	`1`	—
`lora_url`	string	Optional: The URL of a LORA to use	`—`	—
`model`	string	The model to use. 1.3b is faster, but 14b is better quality. A LORA either works with 1.3b or 14b, depending on the version it was trained on.	`"14b"`	1.3b14b
`negative_prompt`	string	Things you do not want to see in your video	`""`	—
`resolution`	string	The resolution of the video. 720p is not supported for 1.3b.	`"480p"`	480p720p
`sample_guide_scale`	number	Higher guide scale makes prompt adherence better, but can reduce variation	`5`	min: 0, max: 10
`sample_shift`	number	Sample shift factor	`8`	min: 0, max: 10
`sample_steps`	integer	Number of generation steps. Fewer steps means faster generation, at the expensive of output quality. 30 steps is sufficient for most prompts	`30`	min: 1, max: 60
`seed`	integer	Set a seed for reproducibility. Random by default.	`—`	—

promptrequiredstring

Text prompt for video generation

aspect_ratiostring

The aspect ratio of the video. 16:9, 9:16, 1:1, etc.

Default: "16:9"

16:99:161:1

fast_modestring

Speed up generation with different levels of acceleration. Faster modes may degrade quality somewhat. The speedup is dependent on the content, so different videos may see different speedups.

Default: "Balanced"

OffBalancedFast

framesinteger

The number of frames to generate (1 to 5 seconds)

Default: 81

1733496581

imagestring

Image to use as a starting frame for image to video generation.

lora_strength_clipnumber

Strength of the LORA applied to the CLIP model. 0.0 is no LORA.

Default: 1

lora_strength_modelnumber

Strength of the LORA applied to the model. 0.0 is no LORA.

Default: 1

lora_urlstring

Optional: The URL of a LORA to use

modelstring

The model to use. 1.3b is faster, but 14b is better quality. A LORA either works with 1.3b or 14b, depending on the version it was trained on.

Default: "14b"

1.3b14b

negative_promptstring

Things you do not want to see in your video

Default: ""

resolutionstring

The resolution of the video. 720p is not supported for 1.3b.

Default: "480p"

480p720p

sample_guide_scalenumber

Higher guide scale makes prompt adherence better, but can reduce variation

Default: 5min: 0, max: 10

sample_shiftnumber

Sample shift factor

Default: 8min: 0, max: 10

sample_stepsinteger

Number of generation steps. Fewer steps means faster generation, at the expensive of output quality. 30 steps is sufficient for most prompts

Default: 30min: 1, max: 60

seedinteger

Set a seed for reproducibility. Random by default.

Version: c48fa8ec65b1Updated: 7/25/202641.3K runs