infery
← All models

LTX-2.3 22B

ltx-2.3-22b-extend-video-lora

Video generationby Lightricks

Extend video with audio using LTX-2.3 and custom LoRA

Example

Details

Accepts
video

Pricing

Price
0.226 cr / megapixel

Prices in credits (1 credit = $0.01).

Data schema

Input

FieldTypeDescription
fpsnumberThe frames per second of the generated video.
seedThe seed for the random number generator.
lorasrequiredarrayThe LoRAs to use for the generation.
promptrequiredstringThe prompt to generate the video from.
schedulerstringenum: ltx2, linear_quadratic, betaThe scheduler to use.
sync_modebooleanIf `True`, the media will be returned as a data URI and the output data won't be available in the request history.
video_urlrequiredstringThe URL of the video to extend.
num_framesintegerThe number of frames to generate.
video_sizeThe size of the generated video.
camera_lorastringenum: dolly_in, dolly_out, dolly_left, dolly_right, jib_up, jib_downThe camera LoRA to use. This allows you to control the camera movement of the generated video more accurately than just prompting the model to move the camera.
accelerationstringenum: none, regular, high, fullThe acceleration level to use.
end_image_urlThe URL of the image to use as the end of the extended video.
video_qualitystringenum: low, medium, high, maximumThe quality of the generated video.
audio_strengthnumberAudio conditioning strength. Lower values represent more freedom given to the model to change the audio content.
generate_audiobooleanWhether to generate audio for the video.
use_multiscalebooleanWhether to use multi-scale generation. If True, the model will generate the video at a smaller scale first, then use the smaller video to guide the generation of a video at or above your requested size. This results in better coherence and details.
video_strengthnumberVideo conditioning strength. Lower values represent more freedom given to the model to change the video content.
audio_cfg_scalenumberThe Classifier-Free Guidance (CFG) scale for the audio. Higher values result in more consistent and focused audio content.
audio_stg_scalenumberThe Spatiotemporal Guidance (STG) scale for the audio. Higher values result in more consistent and focused audio content.
match_input_fpsbooleanWhen true, match the output FPS to the input video's FPS instead of using the default target FPS.
negative_promptstringThe negative prompt to generate the video from.
video_cfg_scalenumberThe Classifier-Free Guidance (CFG) scale for the video. Higher values result in more consistent and focused video content.
video_stg_scalenumberThe Spatiotemporal Guidance (STG) scale for the video. Higher values result in more consistent and focused video content.
extend_directionstringenum: forward, backwardDirection to extend the video. 'forward' extends from the end of the video, 'backward' extends from the beginning.
video_write_modestringenum: fast, balanced, smallThe write mode of the generated video.
camera_lora_scalenumberThe scale of the camera LoRA to use. This allows you to control the camera movement of the generated video more accurately than just prompting the model to move the camera.
video_output_typestringenum: X264 (.mp4), VP9 (.webm), PRORES4444 (.mov), GIF (.gif)The output type of the generated video.
end_image_strengthnumberThe strength of the end image to use for the video generation.
num_context_framesintegerThe number of frames to use as context for the extension.
num_inference_stepsintegerThe number of inference steps to use.
audio_modality_scalenumberThe modality scale for the audio. Controls the ratio between video and audio modalities.
use_restart_samplingbooleanWhether to use restart sampling. This will inject a small amount of noise during each denoising step, which can help improve the quality of the generated video.
video_modality_scalenumberThe modality scale for the video. Controls the ratio between video and audio modalities.
audio_rescaling_scalenumberThe rescaling scale for the audio. Controls the ratio between classifier-free guidance and spatiotemporal guidance.
enable_safety_checkerbooleanWhether to enable the safety checker.
video_rescaling_scalenumberThe rescaling scale for the video. Controls the ratio between classifier-free guidance and spatiotemporal guidance.
enable_prompt_expansionbooleanWhether to enable prompt expansion.
gradient_estimation_gammanumberThe gamma of gradient estimation during denoising. Set to 0 to disable.
distill_lora_first_pass_scalenumberThe scale of the distill LoRA to use for the first pass. Set to 0 to disable.
distill_lora_second_pass_scalenumberThe scale of the distill LoRA to use for the second and subsequent passes.

Output

FieldTypeDescription
seedintegerThe seed used for the random number generator.
videoThe generated video.
promptstringThe prompt used for the generation.