Appearance

LTX-2.3 22B Video Trainer

POST/v1/run

Ltx23 Video Trainer by Lightricks - advanced AI model for training. Delivers high-quality results with fast inference, suitable for both creative and production workflows.

Request body

Submit an async generation request. The model field selects the model; other fields are model-specific input parameters.

stringmodelrequired

Model identifier. Set to lightricks/ltx23-video-trainer.

Default: lightricks/ltx23-video-trainer

Optional<string>aspect_ratio

The aspect ratio of the generated image.

Allowed values: 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16

Optional<string>resolution

Resolution to use for training. Higher resolutions require more memory.

Allowed values: low, medium, high

Default: medium

Optional<number>learning_rate

Learning rate for optimization. Higher values can lead to faster training but may cause overfitting.

Range: 0.000001 to 1

Default: 0.0002

Optional<integer>frame_rate

Target frames per second for the video.

Range: 8 to 60

Default: 25

Optional<array<object>>validation

A list of validation prompts to use during training. When providing an image, _all_ validation inputs must have an image.

Default:

Optional<integer>rank

The rank of the LoRA adaptation. Higher values increase capacity but use more memory.

Allowed values: 8, 16, 32, 64, 128

Default: 32

Optional<boolean>audio_preserve_pitch

When audio duration doesn't match video duration, stretch/compress audio without changing pitch. If disabled, audio is trimmed or padded with silence.

Default: true

Optional<integer>number_of_frames

Number of frames per training sample. Must satisfy frames % 8 == 1 (e.g., 1, 9, 17, 25, 33, 41, 49, 57, 65, 73, 81, 89, 97).

Range: 9 to 121

Default: 89

Optional<integer>number_of_steps

The number of training steps.

Range: 100 to 20000

Default: 2000

Optional<string>validation_aspect_ratio

The aspect ratio to use for validation.

Allowed values: 16:9, 1:1, 9:16

Default: 1:1

Optional<boolean>with_audio

Enable joint audio-video training. If None (default), automatically detects whether input videos have audio. Set to True to force audio training, or False to disable.

Optional<number>stg_scale

STG (Spatio-Temporal Guidance) scale. 0.0 disables STG. Recommended value is 1.0.

Range: 0 to 3

Default: 1

Optional<number>split_input_duration_threshold

The duration threshold in seconds. If a video is longer than this, it will be split into scenes.

Range: 1 to 60

Default: 30

Optional<boolean>split_input_into_scenes

If true, videos above a certain duration threshold will be split into scenes.

Default: true

Optional<integer>validation_number_of_frames

The number of frames in validation videos.

Range: 9 to 121

Default: 89

Optional<string>validation_negative_prompt

A negative prompt to use for validation.

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

Optional<integer>validation_frame_rate

Target frames per second for validation videos.

Range: 8 to 60

Default: 25

Optional<boolean>generate_audio_in_validation

Whether to generate audio in validation samples.

Default: true

Optional<string>training_data_url

URL to zip archive with videos or images. Try to use at least 10 files, although more is better. **Supported video formats:** .mp4, .mov, .avi, .mkv **Supported image formats:** .png, .jpg, .jpeg Note: The dataset must contain ONLY videos OR ONLY images - mixed datasets are not supported. The archive can also contain text files with captions. Each text file should have the same name as the media file it corresponds to.

Optional<number>first_frame_conditioning_p

Probability of conditioning on the first frame during training. Higher values improve image-to-video performance.

Range: 0 to 1

Default: 0.5

Optional<boolean>audio_normalize

Normalize audio peak amplitude to a consistent level. Recommended for consistent audio levels across the dataset.

Default: true

Optional<string>trigger_phrase

A phrase that will trigger the LoRA style. Will be prepended to captions during training.

Default:

Optional<boolean>debug_dataset

When enabled, the trainer returns a downloadable archive of your preprocessed training data for manual inspection. Use this to verify that your videos, images, and captions were processed correctly before committing to a full training run.

Default: false

Optional<boolean>auto_scale_input

If true, videos will be automatically scaled to the target frame count and fps. This option has no effect on image datasets.

Default: false

Optional<string>validation_resolution

The resolution to use for validation.

Allowed values: low, medium, high

Default: high

Response Schema

The submit endpoint returns an accepted generation task. Poll the result endpoint with the returned id for terminal outputs or errors.

Optional<string>error

Error message if the task failed. Empty on success.

stringidrequired

Unique identifier for the generation task.

Optional<string>model

Model ID used for the prediction.

Optional<array>outputs

Array of generated content. Empty when status is not completed.

stringstatusrequired

Status of the task: pending, running, completed, failed, or timeout.

Allowed values: pending, running, completed, failed, timeout

Model capabilities

array<string>capability_tagsrequired

Capabilities declared by the model registry.

Default: training

stringexecution_moderequired

Execution mode declared by the model registry.

Default: async