lightricks/ltx-video-13b-dev
Ltx Video 13b Dev is Lightricks's text-to-video AI model. Turn written scripts and prompts into professional-quality video clips with realistic motion, lighting, and scene composition.
Example output — click Run to generate your own
Send your first request
OpenAI-compatible endpoint with unified authentication and usage tracking.
https://api.sandbase.ai/v1/runlightricks/ltx-video-13b-devInput Schema
15 parameters · 1 required · 14 optional
| Parameter | Type | Required | Description |
|---|---|---|---|
prompt | string | Required | Text prompt to guide generation |
seed | integer | Optional | Random seed for generation |
loras | object[] | Optional | LoRA weights to use for generation · Default: [] |
frame_rate | integer | Optional | The frame rate of the video. · Min: 1 · Max: 60 · Default: 24 |
num_frames | integer | Optional | The number of frames in the video. · Min: 9 · Max: 1441 · Default: 121 |
resolution | string | Optional | Resolution of the generated video. · Options: 480p, 720p · Default: "720p" 480p720p |
aspect_ratio | string | Optional | The aspect ratio of the generated image. · Options: 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 21:916:93:24:35:41:14:53:42:39:16 |
expand_prompt | boolean | Optional | Whether to expand the prompt using a language model. · Default: false |
reverse_video | boolean | Optional | Whether to reverse the video. · Default: false |
enable_detail_pass | boolean | Optional | Whether to use a detail pass. If True, the model will perform a second pass to refine the video and enhance details. This incurs a 2.0x cost multiplier on the base price. · Default: false |
temporal_adain_factor | number | Optional | The factor for adaptive instance normalization (AdaIN) applied to generated video chunks after the first. This can help deal with a gradual increase in saturation/contrast in the generated video by normalizing the color distribution across the video. A high value will ensure the color distribution is more consistent across the video, while a low value will allow for more variation in color distribution. · Min: 0 · Max: 1 · Default: 0.5 |
tone_map_compression_ratio | number | Optional | The compression ratio for tone mapping. This is used to compress the dynamic range of the video to improve visual quality. A value of 0.0 means no compression, while a value of 1.0 means maximum compression. · Min: 0 · Max: 1 · Default: 0 |
first_pass_num_inference_steps | integer | Optional | Number of inference steps during the first pass. · Min: 2 · Max: 50 · Default: 30 |
second_pass_skip_initial_steps | integer | Optional | The number of inference steps to skip in the initial steps of the second pass. By skipping some steps at the beginning, the second pass can focus on smaller details instead of larger changes. · Min: 0 · Max: 49 · Default: 17 |
second_pass_num_inference_steps | integer | Optional | Number of inference steps during the second pass. · Min: 2 · Max: 50 · Default: 30 |
Output Schema
| Field | Type | Description |
|---|---|---|
id | string | Unique identifier for the generation task |
status | string | Task status: pending, running, completed, failed, timeout |
model | string | Model used for the generation |
outputs | array | Array of output items |
outputs[].url | string | URL of the generated artifact |
outputs[].content_type | string | MIME type (e.g. image/png, video/mp4) |
error | object | null | Error details if failed, null on success |
error.type | string | Machine-readable error type code |
error.message | string | Human-readable error description |
Async Workflow
This model uses asynchronous execution. Submit a request and poll for the result.
- Submit — POST to /v1/run, receive an
id - Poll — GET /v1/run/{id} until status is
completed,failed, ortimeout - Retrieve — Read
outputsfrom the completed response
Code Examples
Ready-to-run snippets
# Step 1: Submit
curl -X POST https://api.sandbase.ai/v1/run \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "lightricks/ltx-video-13b-dev",
"loras": [],
"prompt": "A cinematic fast-tracking shot follows a vintage, teal camper van as it descends a winding mountain trail. The van, slightly weathered but well-maintained, is the central focus, its retro design emphasized by the motion blur. Medium shot reveals the dusty, ochre trail, edged with vibrant green pine trees. Close-up on the van's tires shows the gravel spraying, highlighting the speed and rugged terrain. Sunlight filters through the trees, casting dappled shadows on the van and the trail. The background is a hazy, majestic mountain range bathed in warm, golden light. The overall mood is adventurous and exhilarating. High resolution 4k movie scene.",
"frame_rate": 24,
"num_frames": 121,
"resolution": "720p",
"expand_prompt": false,
"reverse_video": false,
"enable_detail_pass": false,
"temporal_adain_factor": 0.5,
"tone_map_compression_ratio": 0,
"first_pass_num_inference_steps": 30,
"second_pass_skip_initial_steps": 17,
"second_pass_num_inference_steps": 30
}'
# Step 2: Poll result (replace <id>)
curl https://api.sandbase.ai/v1/run/<id> \
-H "Authorization: Bearer YOUR_API_KEY"API README
LTX Video-0.9.7 13B
LTX Video-0.9.7 13B is a LTX route built for generative media creation. It turns a focused creative brief and the route-specific inputs into a production-ready media result, giving creators a focused endpoint instead of forcing one generic workflow across materially different production tasks. The route belongs to a family known for audio-visual timing, cinematic motion, and shot-level controllability, so it is best evaluated as a creative system for intentional shots and assets rather than as a one-click novelty generator.
Use this endpoint when the input contract and deliverable match that job exactly. Its documented controls include seed (Random seed for generation); loras (LoRA weights to use for generation); frame_rate (The frame rate of the video); num_frames (The number of frames in the video). Together, these controls help teams plan predictable iterations, compare outputs under stable settings, and connect generation to iterative film, advertising, and social-video workflows without hiding the operational choices that shape the result.
Highlights
- Purpose-built Generative media creation. The route accepts a focused creative brief and the route-specific inputs and produces a production-ready media result; its interface is scoped to that transformation, keeping source assets and creative intent explicit.
- Creative direction. Prompts can describe subject behavior, composition, camera intent, lighting, material, atmosphere, and temporal progression so the result is driven by a shot plan rather than isolated keywords.
- Route-specific control. The request exposes seed (Random seed for generation); loras (LoRA weights to use for generation); frame_rate (The frame rate of the video); num_frames (The number of frames in the video), allowing the same concept to be tested systematically while preserving a repeatable production setup.
- Pipeline-ready output. The generated media asset is returned through the documented asynchronous output contract, which suits review queues, batch iteration, and downstream automation. Editors can review pacing, continuity, lens language, choreography, transitions, temporal artifacts, soundtrack alignment, color response, delivery framing, and cut compatibility before approval.
Pricing
| Configuration | Price |
|---|---|
| Standard request | $0.200000 |
When to Use
| Scenario | Why this model fits |
|---|---|
| Create the exact route output | Choose it when you need generative media creation and already have a focused creative brief and the route-specific inputs. |
| Develop controlled variations | Keep the main brief fixed while changing one documented setting at a time to compare motion, framing, quality, or asset behavior. |
| Build repeatable batches | Use a consistent request shape for catalog, campaign, storyboard, game-asset, or social-content production. |
| Preserve source intent | Prefer this route when the supplied reference material must remain the foundation of a production-ready media result. |
| Connect a media pipeline | Use asynchronous results in an automated review, approval, post-production, or asset-management workflow. |
Prompt Guide
Start with the desired result, then describe the source relationship, subject action, composition or camera behavior, lighting, style, and timing. For generative media creation, state what must remain stable as clearly as what should change. Use only fields exposed by the schema; the example below is structurally valid for this route.
{
"prompt": "A cinematic fast-tracking shot follows a vintage, teal camper van as it descends a winding mountain trail. The van, slightly weathered but well-maintained, is the central focus, its retro design emphasized by the motion blur. Medium shot reveals the dusty, ochre trail, edged with vibrant green pine trees. Close-up on the van's tires shows the gravel spraying, highlighting the speed and rugged terrain. Sunlight filters through the trees, casting dappled shadows on the van and the trail. The background is a hazy, majestic mountain range bathed in warm, golden light. The overall mood is adventurous and exhilarating. High resolution 4k movie scene.",
"seed": 1,
"loras": [],
"frame_rate": 24,
"num_frames": 121
}
Technical Specs
| Specification | Value |
|---|---|
| Model ID | lightricks/ltx-video-13b-dev |
| Workflow | Generative media creation |
| Required inputs | prompt |
seed | integer |
loras | array |
prompt | string |
frame_rate | integer; minimum: 1; maximum: 60 |
num_frames | integer; minimum: 9; maximum: 1441 |
resolution | string; options: 480p, 720p |
aspect_ratio | string; options: 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 |
expand_prompt | boolean |
reverse_video | boolean |
enable_detail_pass | boolean |

