lightricks/ltx-video-0.9.7-distilled

Faster slight quality reduction compared to LTX-Video 13b

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	Aspect ratio for the output video.	`"match_input_image"`	16:91:19:16match_input_image
`conditioning_frames`	integer	Number of frames to use for video-to-video conditioning (only used when video input is provided).	`21`	min: 1, max: 50
`denoise_strength`	number	Denoising strength for final refinement step.	`0.4`	min: 0, max: 1
`downscale_factor`	number	Factor to downscale initial generation (recommended: 2/3 for better quality).	`0.667`	min: 0.1, max: 1
`final_inference_steps`	integer	Number of inference steps for final denoising.	`10`	min: 1, max: 50
`fps`	integer	Frames per second for the output video.	`24`	min: 1, max: 60
`go_fast`	boolean	Enable fast mode with skip layer strategies (20-40% faster but slightly lower quality).	`true`	—
`guidance_scale`	number	Guidance scale. Recommended range: 3.0-3.5.	`3`	min: 1, max: 10
`image`	string(uri)	Input image for image-to-video generation. If provided, will generate video from this image.	`—`	—
`negative_prompt`	string	Negative prompt for video generation.	`"worst quality, inconsistent motion, blurry, jittery, distorted"`	—
`num_frames`	integer	Number of frames to generate.	`121`	min: 9, max: 257
`num_inference_steps`	integer	Number of denoising steps for initial generation.	`24`	min: 2, max: 50
`resolution`	integer	Resolution for the output video (height in pixels).	`720`	480720
`seed`	integer	Random seed for reproducible results. Leave blank for a random seed.	`—`	—
`video`	string(uri)	Input video for video-to-video generation. If provided, will generate video from this video. Takes precedence over image if both are provided.	`—`	—

promptrequiredstring

Text prompt for video generation

aspect_ratiostring

Aspect ratio for the output video.

Default: "match_input_image"

16:91:19:16match_input_image

conditioning_framesinteger

Number of frames to use for video-to-video conditioning (only used when video input is provided).

Default: 21min: 1, max: 50

denoise_strengthnumber

Denoising strength for final refinement step.

Default: 0.4min: 0, max: 1

downscale_factornumber

Factor to downscale initial generation (recommended: 2/3 for better quality).

Default: 0.667min: 0.1, max: 1

final_inference_stepsinteger

Number of inference steps for final denoising.

Default: 10min: 1, max: 50

fpsinteger

Frames per second for the output video.

Default: 24min: 1, max: 60

go_fastboolean

Enable fast mode with skip layer strategies (20-40% faster but slightly lower quality).

Default: true

guidance_scalenumber

Guidance scale. Recommended range: 3.0-3.5.

Default: 3min: 1, max: 10

imagestring

Input image for image-to-video generation. If provided, will generate video from this image.

negative_promptstring

Negative prompt for video generation.

Default: "worst quality, inconsistent motion, blurry, jittery, distorted"

num_framesinteger

Number of frames to generate.

Default: 121min: 9, max: 257

num_inference_stepsinteger

Number of denoising steps for initial generation.

Default: 24min: 2, max: 50

resolutioninteger

Resolution for the output video (height in pixels).

Default: 720

480720

seedinteger

Random seed for reproducible results. Leave blank for a random seed.

videostring

Input video for video-to-video generation. If provided, will generate video from this video. Takes precedence over image if both are provided.

Version: e7f2778ec419Updated: 7/25/20269.3K runs