GET
/v1/schemas/fal/video/fal-ai/ltx-2.3-quality/reference-video-to-video/lora
Accept:
| field | recorded source |
|---|---|
| schema | Source linked above.Source trace07c6405caf9a787a4e5bdda67fd5b03f5195800da55f9523dfa0c917d5f8f781{
"fetchedAt": 1783331098,
"extractorVersion": "1",
"verifiedAt": null
} |
Request body
24 properties · 3 required · 2 $defs| property | type | description | constraints |
|---|---|---|---|
| prompt* | string | The prompt to guide generation. | |
| video_url* | string | The URL of the reference video that supplies motion/structure. | |
| loras* | LoRAInput[] | Up to 3 LoRAs to apply on top of LTX-2.3. Each path is downloaded through the registry SSRF-safe downloader before it is loaded. Max size: 3 GB per LoRA. | items ≤ 3 |
| num_frames | integer | The number of output frames to generate. This IC-LoRA endpoint caps the resolution x frames volume for stability (~150 frames at 720p-class; more at smaller sizes). A request above the budget returns a 422 — reduce num_frames or use a smaller resolution. | 9 ≤ n ≤ 481 · default: 121 |
| guidance_scale | number | Classifier-free guidance scale. The default is tuned for fast, high-quality generation. | 1 ≤ n ≤ 20 · default: 1 |
| frames_per_second | number | Frames per second of the generated video. | 1 ≤ n ≤ 60 · default: 24 |
| preserve_original_video | boolean | True video-to-video: the base video_url is used as the generation's starting point, while control_video_url still drives the structure. The amount of the original preserved is controlled by `strength` (denoise): lower `strength` keeps more of the original vide… | default: false |
| enable_prompt_expansion | boolean | Whether to enable prompt expansion. | default: true |
| enable_safety_checker | boolean | Whether to enable the safety checker. | default: true |
| sync_mode | boolean | If True, the media is returned as a data URI inline in the response. Useful for short-lived requests and tests. | default: false |
| num_inference_steps | integer | Number of inference steps. Defaults to 15 and can be increased up to 30. | 8 ≤ n ≤ 30 · default: 15 |
| seed | integer | Random seed for reproducibility. If None, a random seed is chosen. | nullable |
| strength | number | Sampler denoise strength for reference-video-to-video. With preserve_original_video on, this is the video-to-video amount: lower values keep more of the original video's pixels (e.g. 0.5 = keep ~50%), 1.0 fully regenerates. Without preserve_original_video it o… | 0 ≤ n ≤ 1 · default: 1 |
| skip_control_preprocess | boolean | Skip the built-in control estimation (depth / edge / pose) and use control_video_url directly as the control signal. Requires control_video_url; ignored if it is not set. With resolution='auto', the output follows the source video size/aspect up to the LTX lim… | default: false |
| video_strength | number | Video conditioning strength. Lower values give the model more freedom to change the reference video motion/structure. | 0 ≤ n ≤ 1 · default: 0.6 |
| generate_audio | boolean | Whether to include audio in the returned video. When disabled, the final MP4 is returned without an audio track. | default: true |
| video_quality | enum | The quality preset of the generated video. | low | medium | high | maximum · default: "high" |
| negative_prompt | string | The negative prompt to steer generation away from. | default: "color distortion, overexposure,… |
| resolution | ImageSize | enum | The size of the generated video. In direct-control mode (skip_control_preprocess with control_video_url), 'auto' follows the source video size/aspect up to the LTX limits and explicit sizes are honored with 64px alignment. In the built-in control estimation mo… | default: "auto" |
| mid_image_url | string | Optional middle-frame keyframe image. When set, the look/appearance is anchored to this image around the middle of the clip (in addition to image_url at the start), so the appearance follows a moving camera instead of drifting. Requires image_url. | nullable |
| control_video_url | string | Optional pre-computed control video (e.g. an already-rendered depth / edge / pose composite). When provided together with skip_control_preprocess, the built-in control estimation is skipped and this video is used directly as the control signal, resampled to the output resolution and frame count. | nullable |
| video_write_mode | enum | The write mode of the generated video. | fast | balanced | small · default: "balanced" |
| image_url | string | Optional reference image for style/character anchoring. This is the FIRST-frame keyframe. | nullable |
| end_image_url | string | Optional last-frame keyframe image. When set, the look/appearance is anchored to this image at the end of the clip. Useful for shots where the camera travels to a different place than the start image. Requires image_url. | nullable |
Types
2 $defsImageSizeobject
| height | integer | The height of the generated image. | 0 < n ≤ 14142 · default: 512 |
| width | integer | The width of the generated image. | 0 < n ≤ 14142 · default: 512 |
LoRAInputobject
A single LoRA to apply on top of the LTX-2.3 base model.
| scale | number | Strength of the LoRA application (0.0 to 2.0). Higher values = stronger effect on the generated video. | 0 ≤ n ≤ 2 · default: 1 |
| path* | string | Direct HTTP(S) download URL of the LoRA file (.safetensors). The file is downloaded through an SSRF-safe downloader before it is loaded. Max size: 3 GB. | length ≤ 2048 |
| transformer | enum | Forward-compatibility flag for multi-transformer architectures. LTX-2.3 has a single transformer, so 'both' is the right choice for almost all users. | high | low | both · default: "both" |
Validate a payload
before you spend tokensshell
$ curl -X POST https://modelschemas.com/v1/validate \
-d '{"provider":"fal","endpointId":"fal-ai/ltx-2.3-quality/reference-video-to-video/lora","payload":{…}}'
→ {"valid": false, "errors": [{"path": "#", "keyword": "required", …}]}