LTX 2.5 API docs (Fast / Pro)
LightricksThe LTX 2.5 API on apimodels.app is one API key and one endpoint for Lightricks’ LTX-2.5 video model: it turns a prompt, one first-frame image or an audio track into a 6–20 second clip with sound. There are two model ids: ltx-2.5-fast (720p / 1080p / 1440p; text, image and audio to video) and ltx-2.5-pro (720p / 1080p; text and image to video). Billing is per second of output, from $0.081/s on Fast and $0.108/s on Pro, 10% below Lightricks’ list price, and failed tasks are not charged.
Last updated: 2026-11-20
Models and prices
LTX 2.5 has two model ids, and the price depends only on resolution and output seconds; mode and frame rate do not change the rate.
| Model id | 720p | 1080p | 1440p | Lightricks list price |
|---|---|---|---|---|
| ltx-2.5-fast | $0.081/s | $0.117/s | $0.171/s | $0.09 / $0.13 / $0.19 |
| ltx-2.5-pro | $0.108/s | $0.153/s | Not available here | $0.12 / $0.17 |
- Examples: Fast 720p 6 s $0.486; Fast 1080p 10 s $1.17; Fast 1080p 20 s $2.34; Fast 1440p 10 s $1.71; Pro 720p 6 s $0.648; Pro 1080p 10 s $1.53.
- Audio-to-video is billed on the output length to 0.01 s: 12.5 s at 1080p is 12.5 × $0.117 = $1.4625. If duration is omitted and the audio length cannot be read, the task reserves the maximum (20 s, or 10 s) and settles on the delivered length.
- Failed tasks are not charged. Result files are kept for 30 days.
Five modes, picked from your inputs
LTX 2.5 has no separate image-to-video or audio-to-video model id: a request with audio_url is audio-to-video (Fast only), otherwise a first-frame image makes it image-to-video, otherwise it is text-to-video. A prompt is required in every mode.
| Mode | Model id | Triggered by | Length |
|---|---|---|---|
| Text-to-video | ltx-2.5-fast | prompt only | 6–20 s even; 6 / 8 / 10 at 1440p or 48 / 50 fps |
| Image-to-video | ltx-2.5-fast | prompt + image_url (first frame) | same; aspect ratio follows the image |
| Audio-to-video | ltx-2.5-fast | prompt + audio_url (image_url optional) | follows the audio, or duration 2–20 s (≤10 at 1440p or 48 / 50 fps) |
| Text-to-video | ltx-2.5-pro | prompt only | 6 / 8 / 10 s |
| Image-to-video | ltx-2.5-pro | prompt + image_url (first frame) | 6 / 8 / 10 s; aspect ratio follows the image |
To pin the mode, use a suffixed alias: ltx-2.5-fast-text-to-video, ltx-2.5-fast-image-to-video, ltx-2.5-fast-audio-to-video, ltx-2.5-pro-text-to-video, ltx-2.5-pro-image-to-video (the dotless ltx-2-5-… spellings work too). If the inputs do not match the alias, the request returns a 400 instead of running another mode.
Create a task and poll for the result
POST /v1/video/generations is asynchronous: the create call returns a taskId, then you poll GET /v1/video/generations?task_id= until state is completed and read data.resultUrls[0]; or pass callback_url and we call your URL when the task ends.
curl -X POST https://api.apimodels.app/v1/video/generations \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "ltx-2.5-fast",
"prompt": "Close-up of an old fisherman mending a net on a wooden pier at dawn. He pulls the cord tight, looks up and says, \"Storm is coming.\" The camera slowly pushes in while gulls cry.",
"resolution": "1080p",
"fps": 24,
"duration": 10,
"aspect_ratio": "16:9"
}'
# Poll until state is "completed" (or "failed")
curl "https://api.apimodels.app/v1/video/generations?task_id=TASK_ID" \
-H "Authorization: Bearer YOUR_API_KEY"state has three values: pending (queued or generating), completed (success) and failed (terminal, not charged).
curl examples for all five modes
Each example is copy-paste ready; replace YOUR_API_KEY and the media URLs with your own.
1. Fast text-to-video
curl -X POST https://api.apimodels.app/v1/video/generations \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "ltx-2.5-fast",
"prompt": "Close-up of an old fisherman mending a net on a wooden pier at dawn. He pulls the cord tight, looks up and says, \"Storm is coming.\" The camera slowly pushes in while gulls cry.",
"resolution": "1080p",
"fps": 24,
"duration": 10,
"aspect_ratio": "16:9"
}'2. Fast image-to-video (first frame)
curl -X POST https://api.apimodels.app/v1/video/generations \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "ltx-2.5-fast",
"prompt": "The woman in the photo turns toward the window and smiles as curtains move in the breeze; soft street noise outside.",
"image_url": "https://your-domain.com/first-frame.jpg",
"resolution": "1080p",
"fps": 24,
"duration": 6
}'3. Fast audio-to-video (lip sync / talking head)
curl -X POST https://api.apimodels.app/v1/video/generations \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "ltx-2.5-fast",
"prompt": "A young presenter speaks to the camera in a bright studio, natural hand gestures, medium shot.",
"audio_url": "https://your-domain.com/voice.mp3",
"image_url": "https://your-domain.com/portrait.jpg",
"resolution": "1080p",
"fps": 25
}'
# duration left out: the clip follows the audio (max 20 s; 10 s at 1440p or 48/50 fps).
# Add "duration": 8.5 to cut the first 8.5 s instead.4. Pro text-to-video
curl -X POST https://api.apimodels.app/v1/video/generations \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "ltx-2.5-pro",
"prompt": "Wide establishing shot of a misty mountain lake at sunrise, a lone canoe drifting, the camera slowly cranes down; birdsong and still water.",
"resolution": "1080p",
"fps": 24,
"duration": 10,
"aspect_ratio": "3:2",
"seed": 42
}'5. Pro image-to-video
curl -X POST https://api.apimodels.app/v1/video/generations \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "ltx-2.5-pro",
"prompt": "The perfume bottle on the marble counter catches the morning light as the camera slowly arcs to the right; a soft piano note rings out.",
"image_url": "https://your-domain.com/bottle.jpg",
"resolution": "1080p",
"fps": 24,
"duration": 8
}'Request parameters
An LTX 2.5 request takes only the fields below; a last frame, reference images or video input are rejected with a 400, never silently dropped.
| Field | Required | Type | Description |
|---|---|---|---|
| model | Yes | string | "ltx-2.5-fast" or "ltx-2.5-pro" (or an alias above) |
| prompt | Yes | string | 1–5,000 characters. Required in every mode |
| image_url | No | string | One first frame: public URL or base64 data URI, JPG / PNG / WEBP, up to 10 MB. Makes it image-to-video (the opening frame of an audio-to-video clip on Fast). first_frame_url / image also accepted |
| audio_url | No | string | Fast only. MP3 / WAV / M4A / AAC / OGG / FLAC, at least 2 s, up to 50 MB, kept as the soundtrack. Makes it audio-to-video |
| resolution | No | string | Fast: 720p / 1080p / 1440p; Pro: 720p / 1080p. Default 1080p |
| fps | No | integer | 24 / 25 / 48 / 50, default 24, does not change the price |
| duration | No | number | Text / image: Fast 6–20 even (6 / 8 / 10 at 1440p or 48 / 50 fps), Pro 6 / 8 / 10, default 10. Audio-to-video: optional 2–20 s with decimals (≤10 at 1440p or 48 / 50 fps); omitted = follows the audio |
| aspect_ratio | No | string | Text-to-video only: 16:9 / 9:16 / 1:1 / 3:2 / 2:3, default 16:9. Leave it out for image-to-video (follows the image); audio-to-video without an image is always 16:9 |
| seed | No | integer | 0–2147483647; random when omitted |
| callback_url | No | string | Webhook URL called when the task ends |
Resolution, frame rate and length combinations
LTX 2.5 lengths come in fixed steps, and which steps you get depends on the tier, the resolution and the frame rate.
| Model id | Resolution | fps | Text / image length | Audio-to-video max |
|---|---|---|---|---|
| ltx-2.5-fast | 720p / 1080p | 24 / 25 | 6, 8, 10, 12, 14, 16, 18, 20 | 20 s |
| ltx-2.5-fast | 720p / 1080p | 48 / 50 | 6, 8, 10 | 10 s |
| ltx-2.5-fast | 1440p | 24 / 25 / 48 / 50 | 6, 8, 10 | 10 s |
| ltx-2.5-pro | 720p / 1080p | 24 / 25 / 48 / 50 | 6, 8, 10 | Not available |
Errors and failures
An invalid request returns a 400 before any task is created and is not charged; these are the most common LTX 2.5 messages, verbatim.
| Status | When | Returned msg |
|---|---|---|
| 400 | audio_url sent to Pro | audio-to-video (audio_url) is only available on ltx-2.5-fast; ltx-2.5-pro supports text-to-video and image-to-video |
| 400 | 1440p on Pro | resolution must be 720p or 1080p on ltx-2.5-pro (got "1440p"); 1440p is only on ltx-2.5-fast |
| 400 | Length outside the allowed steps | duration must be 6, 8 or 10 at 1440p or 48/50 fps (got "12") |
| 400 | aspect_ratio with a first frame | aspect_ratio must be omitted when a first-frame image is given — the video keeps the image's aspect ratio (got "1:1") |
| 400 | Last frame sent | ltx-2.5-fast has no last-frame input; remove last_frame_url (only a first frame is supported, as image_url) |
| 400 | Audio shorter than duration | audio_url is about 3.2s long, shorter than duration 5s — lower duration or omit it to follow the audio |
| 401 | Missing or invalid API key | — |
| 402 | Balance below the price of the clip | — |
| state: failed | Content review rejected it, or the upstream was busy | Not charged; adjust the input and retry |
Limits
- Not available here: 4K, last frame, reference images, video input, automatic duration and silent output (every clip has sound). Lightricks’ own API offers 4K, a last frame and Pro audio-to-video.
- Pro has no audio-to-video and no 1440p, and stops at 10 seconds.
- Prompts up to 5,000 characters; first frame up to 10 MB; audio at least 2 s and up to 50 MB.
- Success is not 100%: content review or a busy upstream can fail a task, and failed tasks are not charged.
- Lightricks’ prompting guide notes that on-screen text and chaotic physics are still uneven.
Sources
Checked 2026-11-20; prices are apimodels.app prices, and Lightricks’ figures are as published on its own pages.
Related
FAQ
How much does the LTX 2.5 API cost?
Per second of output on apimodels.app: ltx-2.5-fast $0.081 (720p), $0.117 (1080p), $0.171 (1440p); ltx-2.5-pro $0.108 (720p), $0.153 (1080p). That is 10% below Lightricks’ list price. A 10-second 1080p clip is $1.17 on Fast and $1.53 on Pro; audio-to-video is billed on the output length to 0.01 s. Failed tasks are not charged.
How do I do LTX 2.5 image to video through the API?
Add image_url (a public URL or base64 data URI; JPG, PNG or WEBP up to 10 MB) to the same request and leave aspect_ratio out; the clip opens on that image and keeps its shape. Only a first frame is supported, so last_frame_url is rejected with a 400.
How does LTX 2.5 audio to video (lip sync) work?
Send audio_url to ltx-2.5-fast with a prompt, and optionally a portrait as image_url. The audio (MP3, WAV, M4A, AAC, OGG or FLAC, 2 s or longer, up to 50 MB) becomes the soundtrack and the picture follows it. Without duration the clip follows the audio up to 20 s (10 s at 1440p or 48 / 50 fps). ltx-2.5-pro has no audio-to-video here.
What is the max length and max resolution of LTX 2.5?
On apimodels.app: ltx-2.5-fast reaches 20 seconds at 720p or 1080p with 24 / 25 fps, and 1440p at up to 10 seconds; ltx-2.5-pro is 720p or 1080p at 6, 8 or 10 seconds. There is no 4K here, although Lightricks’ own API offers it.
How do I force a specific LTX 2.5 mode?
Use a suffixed alias such as ltx-2.5-fast-image-to-video or ltx-2.5-pro-text-to-video. If the inputs do not match the alias (for example audio_url on an image-to-video alias), the request returns a 400 instead of running a different mode. With the plain ids, the mode follows the inputs.