Lightricks

This page is auto-generated from model configurations. Last updated: 2026-03-13.

This reference lists all available Lightricks video generation models and their parameters. Use these parameter names when calling the Generation API.


LTX-2 19b

High-quality image-to-video and text-to-video with synchronized audio.

Model ID: model_ltx-2-19b

Capabilities: txt2video, img2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-19b/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
promptstringYes----Text prompt for video generation (1-5000 characters)
imagefileNo----Input image for image-to-video (applied at frame 0). Leave empty for text-to-video.
imageStrengthnumberNo101-Image conditioning strength (0-1) for image-to-video.
negativePromptstringNo----What to avoid in generation
durationnumberNo6320-Video duration in seconds. Durations below 6s may produce lower quality.
aspectRatiostringNoauto--auto, 16:9, 9:16, 4:3, 3:4, 1:1, 21:9Video aspect ratio. 'Match Input' detects from input image.
resolutionstringNo1080p--720p, 1080p, 1440p, 2160pVideo resolution. Ignored if width and height are provided.
widthnumberNo-644096-Explicit video width in pixels. Must be used with height. Rounded to multiple of 64.
heightnumberNo-644096-Explicit video height in pixels. Must be used with width. Rounded to multiple of 64.
fpsnumberNo25160-Frame rate in frames per second (1-60)
cameraMotionstringNonone--none, static, dolly_in, dolly_out, dolly_left, dolly_right, jib_up, jib_downCamera movement type in the generated video
cameraMotionStrengthnumberNo102-Strength of camera motion (0-2)
numInferenceStepsnumberNo401100-Number of inference steps. Higher = better quality, slower.
guidanceScalenumberNo3120-Guidance scale. Higher = stronger prompt adherence.
generateAudiobooleanNotrue---Whether to generate synchronized audio
enhancePromptbooleanNofalse---Enhance the prompt for better video generation
seednumberNo-0--Random seed for reproducibility. Leave blank for random.

LTX-2 19b Extend Video

Extend videos by generating continuation frames. Input a video and describe how it should continue.

Model ID: model_ltx-2-19b-extend-video

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-19b-extend-video/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
promptstringYes----Text prompt describing how the video should continue
videofileYes----Input video to extend. The model generates continuation frames based on this video.
videoStrengthnumberNo101-Video conditioning strength. Higher = more faithful to input video motion/style.
extendDurationnumberNo4120-Duration of new video to generate in seconds. Additional duration added after the input video.
contextDurationnumberNo-0.1--Seconds from input video to use as context. Leave empty to use entire video.
aspectRatiostringNoauto--auto, 21:9, 16:9, 4:3, 3:4, 1:1, 9:16Video aspect ratio. 'Match Input' detects from input video.
resolutionstringNoauto--auto, 720p, 1080p, 1440p, 2160pVideo resolution. 'Match Input' replicates input video dimensions (recommended for extend).
widthnumberNo-644096-Explicit video width in pixels. Must be used with height. Rounded to multiple of 64.
heightnumberNo-644096-Explicit video height in pixels. Must be used with width. Rounded to multiple of 64.
fpsnumberNo-160-Output frame rate (1-60 fps). Leave empty to match input video.
numInferenceStepsnumberNo401100-Number of denoising steps (more = higher quality, slower)
guidanceScalenumberNo3120-Classifier-free guidance scale (higher = more prompt adherence)
cameraMotionstringNonone--none, static, dolly_in, dolly_out, dolly_left, dolly_right, jib_up, jib_downCamera movement type in the generated continuation
cameraMotionStrengthnumberNo102-Strength of camera motion
generateAudiobooleanNotrue---Whether to generate synchronized audio
audioStrengthnumberNo101-Audio conditioning strength (0-1). Higher preserves more input audio characteristics.
enhancePromptbooleanNofalse---Enhance the prompt for better video generation
seednumberNo-0--Random seed for reproducibility. Leave blank for random.
negativePromptstringNo----What to avoid in generation

LTX-2 19b Fast

Fast tier (8 steps) for rapid iteration with synchronized audio.

Model ID: model_ltx-2-19b-fast

Capabilities: txt2video, img2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-19b-fast/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
promptstringYes----Text prompt for video generation (1-5000 characters)
imagefileNo----Input image for image-to-video (applied at frame 0). Leave empty for text-to-video.
imageStrengthnumberNo101-Image conditioning strength (0-1) for image-to-video.
durationnumberNo6320-Video duration in seconds (3-20). Durations below 6s may produce lower quality. >10s only at 1080p/25fps.
aspectRatiostringNoauto--auto, 21:9, 16:9, 4:3, 3:4, 1:1, 9:16Video aspect ratio. 'Match Input' detects from input image.
resolutionstringNo1080p--720p, 1080p, 1440p, 2160pVideo resolution. Ignored if width and height are provided.
widthnumberNo-644096-Explicit video width in pixels. Must be used with height. Rounded to multiple of 64.
heightnumberNo-644096-Explicit video height in pixels. Must be used with width. Rounded to multiple of 64.
fpsnumberNo25160-Frame rate in frames per second (1-60)
cameraMotionstringNonone--none, static, dolly_in, dolly_out, dolly_left, dolly_right, jib_up, jib_downCamera movement type in the generated video
cameraMotionStrengthnumberNo102-Strength of camera motion (0-2)
generateAudiobooleanNotrue---Whether to generate synchronized audio
enhancePromptbooleanNofalse---Enhance the prompt for better video generation
seednumberNo-0--Random seed for reproducibility. Leave blank for random.

LTX-2 19b Keyframes to Video

Interpolate between multiple keyframe images to generate smooth video transitions with synchronized audio.

Model ID: model_ltx-2-19b-keyframes-to-video

Capabilities: img2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-19b-keyframes-to-video/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
promptstringYes----Text prompt describing the video content and transitions (1-5000 characters)
keyframesinputs_arrayYes----List of keyframe images with timestamps and strengths. At least 2 keyframes required for interpolation.
negativePromptstringNo----Negative prompt to guide what to avoid in generation
resolutionstringNo1080p--720p, 1080p, 1440p, 2160pVideo resolution. Ignored if width and height are provided.
widthnumberNo-644096-Explicit video width in pixels. Must be used with height. Rounded to multiple of 64.
heightnumberNo-644096-Explicit video height in pixels. Must be used with width. Rounded to multiple of 64.
cameraMotionstringNonone--none, static, dolly_in, dolly_out, dolly_left, dolly_right, jib_up, jib_downCamera movement type in the generated video
cameraMotionStrengthnumberNo102-Strength of camera motion (0-2). Higher values = stronger effect.
durationnumberNo6320-Video duration in seconds (3-20). Durations below 6s may produce lower quality.
aspectRatiostringNoauto--auto, 16:9, 9:16, 4:3, 3:4, 1:1, 21:9Video aspect ratio. 'Match Input' detects from first keyframe image.
fpsnumberNo25160-Frame rate in frames per second (1-60)
numInferenceStepsnumberNo401100-Number of inference steps. Higher = better quality, slower.
guidanceScalenumberNo3120-CFG guidance scale. Higher = stronger prompt adherence.
generateAudiobooleanNotrue---Whether to generate synchronized audio
enhancePromptbooleanNofalse---Use Gemma to enhance the prompt for better video generation
seednumberNo-0--Random seed for reproducibility. Leave blank for random.

LTX-2 19b Video to Video

Transform videos with text prompts. Input a reference video and describe the desired output. Supports IC-LoRA control (canny, depth, pose, detailer) and optional first-frame conditioning.

Model ID: model_ltx-2-19b-video-to-video

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-19b-video-to-video/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
promptstringYes----Text prompt for video generation
videofileYes----Reference video. The model generates output conditioned on this video's structure.
videoStrengthnumberNo101-Reference video conditioning strength (0-1). Higher = more faithful to input video structure.
imagefileNo----Optional first-frame image. Applied at frame 0 for additional conditioning.
imageStrengthnumberNo101-First-frame image conditioning strength (0-1).
negativePromptstringNo----Negative prompt to guide what to avoid in generation
resolutionstringNoauto--auto, 720p, 1080p, 1440p, 2160pVideo resolution. 'Match Input' replicates input video dimensions (recommended).
widthnumberNo-644096-Explicit video width in pixels. Must be used with height. Rounded to multiple of 64.
heightnumberNo-644096-Explicit video height in pixels. Must be used with width. Rounded to multiple of 64.
cameraMotionstringNonone--none, static, dolly_in, dolly_out, dolly_left, dolly_right, jib_up, jib_downCamera movement type in the generated video.
cameraMotionStrengthnumberNo102-Strength of camera motion (0-2). Higher values = stronger effect.
durationnumberNo-320-Output video duration in seconds (3-20). Leave empty to match input video.
aspectRatiostringNoauto--auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16Video aspect ratio. 'Match Input' detects from input video (recommended).
fpsnumberNo-160-Output frame rate (1-60 fps). Leave empty to match input video.
audioSourcestringNooriginal--original, generate, noneAudio for output: copy from input, generate new, or none.
modalitystringNonone--none, canny, depth, pose, detailerIC-LoRA control: canny (edges), depth, pose (skeleton), detailer, or none.
modalityStrengthnumberNo102-Strength of IC-LoRA modality (0-2). Higher values = stronger control.
conditioningVideofileNo----Pre-processed conditioning video (canny/depth/pose). Use with matching modality.
numInferenceStepsnumberNo401100-Number of inference steps. Higher = better quality, slower.
guidanceScalenumberNo3120-Classifier-free guidance scale. Higher = stronger prompt adherence.
enhancePromptbooleanNofalse---Use Gemma to enhance the prompt for better video generation
seednumberNo-0--Random seed for reproducibility. Leave blank for random.

LTX-2 Fast

Ideal for rapid ideation and mobile workflows. Perfect for creators who need instant feedback, real-time previews, or high-throughput content.

Model ID: model_ltx-2-fast

Capabilities: txt2video, img2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-fast/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
promptstringYes----Text prompt describing the video to generate
imagefileNo----Input image for image-to-video generation
resolutionstringNo1080p--1080p, 2k, 4kResolution quality of the generated video
durationnumberNo6--6, 8, 10, 12, 14, 16, 18, 20Duration of the video in seconds
generateAudiobooleanNofalse---Generate audio for the video

LTX-2 Pro

Delivers high visual fidelity with fast turnaround. Great for daily content creation, marketing teams, and iterative creative workflows.

Model ID: model_ltx-2-pro

Capabilities: txt2video, img2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-pro/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
promptstringYes----Text prompt describing the video to generate
imagefileNo----Input image for image-to-video generation
resolutionstringNo1080p--1080p, 2k, 4kResolution quality of the generated video
durationnumberNo6--6, 8, 10Duration of the video in seconds
generateAudiobooleanNofalse---Generate audio for the video

LTX-2 Retake

Model ID: model_ltx-2-retake

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-retake/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
videofileYes----Input video to edit
promptstringYes----Text prompt describing the edit
modestringNoreplace_video--replace_audio, replace_video, replace_audio_and_videoMode of operation
startTimenumberNo00--Start time of the edit in seconds
durationnumberNo6216-Duration of the edited section in seconds

LTX-2.3 Fast

Speed-optimized LTX-2.3 video generation. Generates videos with synchronized audio faster than real-time — ideal for rapid prototyping, mobile workflows, and high-volume production.

Model ID: model_ltx-2-3-fast

Capabilities: txt2video, img2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-3-fast/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
promptstringYes----Text prompt describing the video to generate
startImagefileNo----First frame image for image-to-video generation
endImagefileNo----End frame for interpolation (requires first frame). Video interpolates between first and last frame.
resolutionstringNo1080p--1080p, 2k, 4kVideo resolution. Durations longer than 10 seconds only at 1080p with 24 or 25 FPS.
durationnumberNo6--6, 8, 10, 12, 14, 16, 18, 20Duration of the video in seconds
aspectRatiostringNo16:9--16:9, 9:16Aspect ratio of the generated video
fpsnumberNo25--24, 25, 48, 50Frame rate in frames per second
cameraMotionstringNonone--none, dolly_in, dolly_out, dolly_left, dolly_right, jib_up, jib_down, static, focus_shiftCamera motion effect to apply to the generated video
generateAudiobooleanNotrue---Generate synchronized audio for the video

LTX-2.3 Pro

High-fidelity LTX-2.3 video generation. Produces higher visual quality than LTX 2.3 Fast. Supports text-to-video, image-to-video with optional end frame interpolation.

Model ID: model_ltx-2-3-pro

Capabilities: txt2video, img2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-3-pro/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
promptstringYes----Text prompt describing the video to generate
startImagefileNo----First frame image for image-to-video.
endImagefileNo----Optional end frame for interpolation. Video interpolates between first and last frame.
resolutionstringNo1080p--1080p, 2k, 4kVideo resolution (audio_to_video, retake, extend are 1080p only)
durationnumberNo6--6, 8, 10Duration of the video in seconds
aspectRatiostringNo16:9--16:9, 9:16Aspect ratio of the generated video
fpsnumberNo25--24, 25, 48, 50Frame rate in frames per second
cameraMotionstringNonone--none, dolly_in, dolly_out, dolly_left, dolly_right, jib_up, jib_down, static, focus_shiftCamera motion effect to apply to the generated video
generateAudiobooleanNotrue---Generate synchronized audio for the video

LTX-2.3 Pro Audio to Video

Generate video driven by an audio file. Speech cadence controls pacing, musical energy influences motion. $0.10/sec via fal.ai.

Model ID: model_ltx-2-3-pro-audio-to-video

Capabilities: txt2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-3-pro-audio-to-video/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
audiofileYes----Audio file (2-20 seconds). Drives the video generation.
imagefileNo----Optional image to use as the first frame. If not provided, prompt is required.
promptstringNo----Text description of how the video should be generated. Required if image is not provided.
guidanceScalenumberNo5150-Higher values make output follow the prompt more closely. Default 5 for text-only, 9 with image.

LTX-2.3 Pro Extend Video

Add duration to the beginning or end of a video using LTX 2.3 Pro. Extend existing clips with high-fidelity continuation.

Model ID: model_ltx-2-3-pro-extend-video

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-3-pro-extend-video/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
videofileYes----Input video to extend
promptstringYes----Text prompt describing how the video should continue
durationnumberNo6--6, 8, 10Duration of the video in seconds
extendModestringNoend--start, endWhere to add the new duration

LTX-2.3 Pro Retake

Re-generate a section of an existing video. Replace audio, video, or both with LTX 2.3 Pro. 1080p only.

Model ID: model_ltx-2-3-pro-retake

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_ltx-2-3-pro-retake/markdown

ParameterTypeRequiredDefaultMinMaxAllowed ValuesDescription
videofileYes----Input video to edit
promptstringYes----Text prompt describing the edit
retakeModestringNoreplace_audio_and_video--replace_audio, replace_video, replace_audio_and_videoWhat to replace in the section
startTimenumberNo00--Start time in seconds of section to edit
durationnumberNo22--Duration in seconds of section to edit (≥2s)