LTXLTX 2.3 Outpaint API

Expand a video past its original frame — the source stays put inside a wider canvas and the surrounding region is generated to match.

Prompt

Source Video

Generate audio

Duration

Resolution

Aspect ratio

Billed at this model’s published rate.

Result

Your generations will appear here

Other models from LTX

Same vendor, same API call — the differences are below.

Model	Use case	Price	Resolution	Duration	Added
LTX 2.3 Reframe	Video → Video	from $0.11 / sec	720p · 1080p	—	Sep 2026
LTX 2.5 Fast	Text → Video	from $0.099 / sec	720p · 1080p · 1440p …	6–20s	Sep 2026
LTX 2.5 Pro	Text → Video	from $0.132 / sec	720p · 1080p	6–10s	Sep 2026
LTX 2.3 Audio-to-Video	Audio → Video	from $0.08 / sec	—	—	Mar 2026
LTX 2.3 Extend	Video → Video	from $0.11 / sec	—	2–20s	Mar 2026
LTX 2.3 Fast	Text → Video	from $0.06 / sec	1080p · 1440p · 2160p	6–20s	Mar 2026
LTX 2.3 Pro	Text → Video	from $0.08 / sec	1080p · 1440p · 2160p	6–10s	Mar 2026
LTX 2.3 Retake	Video → Video	from $0.11 / sec	—	2–20s	Mar 2026

Install the SDK

TypeScript targets Node 20+ and ships with types; Python needs 3.10+ and is async. Over REST there is nothing to install: call https://api.picsart.com from any HTTP client.

bash

npm install @picsart/ai-sdk

Authenticate

Send your Picsart API key as a bearer token. The SDK does it for you once the client is created.

import { createClient } from '@picsart/ai-sdk';

const ai = createClient({
  apiKey: process.env.PICSART_API_KEY, // sent as: Authorization: Bearer <key>
  apiUrl: 'https://api.picsart.com',
});

python

import os

from picsart_ai_sdk import create_client

ai = create_client(
    api_key=os.environ["PICSART_API_KEY"],  # sent as: Authorization: Bearer <key>
    api_url="https://api.picsart.com",
)

Make a request

Call the model by its id. Over REST, one endpoint serves every model: POST /workflows/v1/models/execute with the model id and its input in the body. The outer params is the envelope every workflow takes; the inner one is this model's input, with the same names generate() takes.

import { createClient } from '@picsart/ai-sdk';

const ai = createClient({
  apiKey: process.env.PICSART_API_KEY,
  apiUrl: 'https://api.picsart.com',
});

const result = await ai.generate('ltx-v2.3-outpaint', {
  prompt: "A serene mountain lake at golden hour, ultra detailed",
  videoUrl: "https://cdn.example.com/input.mp4",
  aspectRatio: "21:9",
  resolution: "720p",
  duration: 6
});

console.log(result.url); // video URL

python

import asyncio
import os

from picsart_ai_sdk import create_client


async def main() -> None:
    async with create_client(
        api_key=os.environ["PICSART_API_KEY"],
        api_url="https://api.picsart.com",
    ) as ai:
        result = await ai.generate("ltx-v2.3-outpaint", {
            "prompt": "A serene mountain lake at golden hour, ultra detailed",
            "videoUrl": "https://cdn.example.com/input.mp4",
            "aspectRatio": "21:9",
            "resolution": "720p",
            "duration": 6,
        })

        print(result.url)  # video URL


asyncio.run(main())

bash

curl -X POST 'https://api.picsart.com/workflows/v1/models/execute' \
  -H 'Authorization: Bearer $PICSART_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "params": {
    "model": "ltx-v2.3-outpaint",
    "params": {
      "prompt": "A serene mountain lake at golden hour, ultra detailed",
      "videoUrl": "https://cdn.example.com/input.mp4",
      "aspectRatio": "21:9",
      "resolution": "720p",
      "duration": 6
    }
  }
}'

Async (submit & poll)

This model can run longer than the sync limit (~20s). generate() waits for you; to hold on to the job instead, submit it and poll. Over REST, send it to /workflows/v1/models/submit and poll /workflows/v1/models/{taskId}/result until response.status is COMPLETED or FAILED.

import { createClient } from '@picsart/ai-sdk';

const ai = createClient({
  apiKey: process.env.PICSART_API_KEY,
  apiUrl: 'https://api.picsart.com',
});

// Returns the generation id at once — store it to pick the job up later.
const generationId = await ai.submit('ltx-v2.3-outpaint', {
  prompt: "A serene mountain lake at golden hour, ultra detailed",
  videoUrl: "https://cdn.example.com/input.mp4",
  aspectRatio: "21:9",
  resolution: "720p",
  duration: 6
});

// Polls until the job completes; throws if it fails or is cancelled.
const result = await ai.result('ltx-v2.3-outpaint', generationId);

console.log(result.url); // video URL

python

import asyncio
import os

from picsart_ai_sdk import create_client


async def main() -> None:
    async with create_client(
        api_key=os.environ["PICSART_API_KEY"],
        api_url="https://api.picsart.com",
    ) as ai:
        # Returns the generation id at once — store it to pick the job up later.
        generation_id = await ai.submit("ltx-v2.3-outpaint", {
            "prompt": "A serene mountain lake at golden hour, ultra detailed",
            "videoUrl": "https://cdn.example.com/input.mp4",
            "aspectRatio": "21:9",
            "resolution": "720p",
            "duration": 6,
        })

        # Polls until the job completes; raises ApiError if it fails or is cancelled.
        result = await ai.result("ltx-v2.3-outpaint", generation_id)

        print(result.url)  # video URL


asyncio.run(main())

bash

# 1. Submit the job
curl -X POST 'https://api.picsart.com/workflows/v1/models/submit' \
  -H 'Authorization: Bearer $PICSART_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "params": {
    "model": "ltx-v2.3-outpaint",
    "params": {
      "prompt": "A serene mountain lake at golden hour, ultra detailed",
      "videoUrl": "https://cdn.example.com/input.mp4",
      "aspectRatio": "21:9",
      "resolution": "720p",
      "duration": 6
    }
  }
}'
# → { "status": "success", "response": { "id": "task_abc123" } }

# 2. Poll until response.status is COMPLETED
curl 'https://api.picsart.com/workflows/v1/models/task_abc123/result' \
  -H 'Authorization: Bearer $PICSART_API_KEY'

Parameters

17 parameters, the same for the SDK and for the inner params over REST. Required ones must be supplied; the rest fall back to their defaults.

Parameter	Type	Required	Default	Details
`prompt` What to make, in plain language: the subject, style, composition and motion. With an input image or video, what to change.	`text`	yes	`—`	max 5000 chars
`videoUrl` The video to work on, as a URL.	`file (video)`	yes	`—`	—
`negativePrompt` What to keep out of the result, such as text, watermarks or blur.	`text`	no	`—`	—
`aspectRatio` Shape of the output as width:height. An auto or adaptive option follows the input image.	`enum`	no	`21:9`	21:9, 16:9, 4:3, 1:1, 3:4, 9:16, 9:21
`resolution` Output size. Higher costs more and takes longer.	`enum`	no	`720p`	480p, 720p, 1080p
`duration` Length of the output, in seconds.	`enum`	no	`6`	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20
`fps` Frames per second of the video.	`range`	no	`24`	1–60 (step 1)
`sourceScale` Source Scale	`range`	no	`1`	0.25–1 (step 0.05)
`videoStrength` Source Strength	`range`	no	`1`	0–1 (step 0.05)
`cfgScale` How closely the result follows the prompt. Higher is stricter, lower more creative.	`range`	no	`1`	1–20
`numInferenceSteps` Denoising steps. More is slower and more detailed.	`range`	no	`15`	8–30
`videoQuality` Video Quality	`enum`	no	`high`	low, medium, high, maximum
`videoWriteMode` Write Mode	`enum`	no	`balanced`	fast, balanced, small
`enhancePrompt` Let the model rewrite the prompt for a better result before generating.	`boolean`	no	`true`	—
`generateAudio` Whether the video gets a generated soundtrack. Audio can cost more.	`boolean`	no	`true`	—
`enableSafetyChecker` Run the vendor's safety check on the result.	`boolean`	no	`true`	—
`seed` Seed for the random generator. The same seed and inputs repeat a result; leave it out for a random one.	`range`	no	`—`	0–2147483647 (step 1)

Response

generate() resolves once the job completes. A video URL. Over REST the output is at response.result, inside the gateway's status envelope with the task id, its status and the amount charged.

json

{
  "items": [
    {
      "url": "https://cdn.picsart.com/…/result"
    }
  ],
  "model": "ltx-v2.3-outpaint"
}

json

{
  "status": "success",
  "response": {
    "id": "task_abc123",
    "status": "COMPLETED",
    "result": {
      "url": "https://cdn.picsart.com/…/result",
      "items": [
        {
          "url": "https://cdn.picsart.com/…/result"
        }
      ],
      "model": "ltx-v2.3-outpaint"
    }
  }
}

Overview

Alongside the unified model APIs, we expose compatibility APIs that take each vendor's original parameters exactly as the vendor defines them — nothing renamed, nothing reshaped on the way through.

That makes them the shortest path onto Picsart if you already work with the vendor directly: the request bodies you've already written keep working as they are. You keep their parameter names and defaults, and you get the vendor's full parameter set rather than the subset that is shared across every model.

The cost is that a request is written for one vendor — switching models later means rewriting it, and results come back in the vendor's own shape. When you'd rather write once and change models freely, use the unified API.

Make a request

Call the workflow with ai.apis.run() in TypeScript or Python, or hit /workflows/{workflow}/execute directly — params is passed through untouched either way.

These endpoints are addressed by workflow name rather than model id: the one in this model's header, since one model runs as one workflow. Authentication is unchanged — your Picsart API key as a bearer token (see Authentication).

import { createClient, ApiRunMode } from '@picsart/ai-sdk';

const ai = createClient({
  apiKey: process.env.PICSART_API_KEY,
  apiUrl: 'https://api.picsart.com',
});

// Calls the 'ltx-2.3-quality/outpaint' workflow directly — params are sent as-is.
const { result, usage } = await ai.apis.run('ltx-2.3-quality/outpaint', {
  prompt: "A cinematic wide shot of the same scene, extending the environment naturally beyond the original frame.",
  video_url: "https://cdn.example.com/input.mp4"
}, {
  mode: ApiRunMode.SYNC,
});

console.log(result); // workflow-specific output
console.log(usage?.credits); // amount charged, in $

python

import asyncio
import os

from picsart_ai_sdk import ApiRunMode, create_client


async def main() -> None:
    async with create_client(
        api_key=os.environ["PICSART_API_KEY"],
        api_url="https://api.picsart.com",
    ) as ai:
        # Calls the 'ltx-2.3-quality/outpaint' workflow directly — params are sent as-is.
        response = await ai.apis.run("ltx-2.3-quality/outpaint", {
            "prompt": "A cinematic wide shot of the same scene, extending the environment naturally beyond the original frame.",
            "video_url": "https://cdn.example.com/input.mp4",
        }, mode=ApiRunMode.SYNC)

        print(response["result"])  # workflow-specific output
        print((response.get("usage") or {}).get("credits"))  # amount charged, in $


asyncio.run(main())

bash

curl -X POST 'https://api.picsart.com/workflows/ltx-2.3-quality/outpaint/execute' \
  -H 'Authorization: Bearer $PICSART_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "params": {
    "prompt": "A cinematic wide shot of the same scene, extending the environment naturally beyond the original frame.",
    "video_url": "https://cdn.example.com/input.mp4"
  }
}'

Async (submit & poll)

This model can run longer than the sync limit (~20s). In the SDK, ai.apis.run() with ApiRunMode.ASYNC polls for you; over HTTP, submit and poll yourself.

import { createClient, ApiRunMode } from '@picsart/ai-sdk';

const ai = createClient({
  apiKey: process.env.PICSART_API_KEY,
  apiUrl: 'https://api.picsart.com',
});

// mode: ASYNC submits the job and polls under the hood — you just await.
const { result } = await ai.apis.run('ltx-2.3-quality/outpaint', {
  prompt: "A cinematic wide shot of the same scene, extending the environment naturally beyond the original frame.",
  video_url: "https://cdn.example.com/input.mp4"
}, {
  mode: ApiRunMode.ASYNC,
});

console.log(result);

python

import asyncio
import os

from picsart_ai_sdk import ApiRunMode, create_client


async def main() -> None:
    async with create_client(
        api_key=os.environ["PICSART_API_KEY"],
        api_url="https://api.picsart.com",
    ) as ai:
        # mode=ASYNC submits the job and polls under the hood — you just await.
        response = await ai.apis.run("ltx-2.3-quality/outpaint", {
            "prompt": "A cinematic wide shot of the same scene, extending the environment naturally beyond the original frame.",
            "video_url": "https://cdn.example.com/input.mp4",
        }, mode=ApiRunMode.ASYNC)

        print(response["result"])


asyncio.run(main())

bash

# 1. Submit the job
curl -X POST 'https://api.picsart.com/workflows/ltx-2.3-quality/outpaint/submit' \
  -H 'Authorization: Bearer $PICSART_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "params": {
    "prompt": "A cinematic wide shot of the same scene, extending the environment naturally beyond the original frame.",
    "video_url": "https://cdn.example.com/input.mp4"
  }
}'
# → { "status": "success", "response": { "id": "wf_abc123" } }

# 2. Poll until response.status is COMPLETED
curl 'https://api.picsart.com/workflows/ltx-2.3-quality/outpaint/wf_abc123/result' \
  -H 'Authorization: Bearer $PICSART_API_KEY'

Parameters

20 parameters, sent inside params. These are the vendor's own names, so they line up one-for-one with the vendor's documentation. Required ones must be supplied; the rest fall back to their defaults.

Parameter	Type	Required	Default	Details
`negative_prompt` The negative prompt to steer generation away from.	`string`	no	`color distortion, overexposure, static, blurry details, subtitles, style, artwork, painting, frame, still, dim overall tone, worst quality, low quality, JPEG compression artifacts, ugly, mutilated, extra fingers, poorly drawn hands, poorly drawn face, deformed, disfigured, malformed limbs, fused fingers, motionless frame, cluttered background, three legs, crowded background, walking backwards`	—
`num_inference_steps` Number of inference steps. Defaults to 15 and can be increased up to 30.	`integer`	no	`15`	8–30
`prompt` The prompt to guide the outpainted video generation.	`string`	yes	`—`	e.g. `A cinematic wide shot of the same scene, extending the environment naturally beyond the original frame.`
`generate_audio` Whether to include audio in the returned video. When disabled, the final MP4 is returned without an audio track.	`boolean`	no	`true`	—
`output_resolution` Output resolution tier. 480p is faster, 720p is the default balance, and 1080p gives a larger canvas when it fits LTX limits.	`string`	no	`720p`	`480p720p1080p`
`guidance_scale` Classifier-free guidance scale. The default is tuned for fast, high-quality generation.	`number`	no	`1`	1–20
`video_strength` Strength of the IC-LoRA video guide for the original content. Higher values preserve the source video more strongly inside the outpainted canvas.	`number`	no	`1`	0–1
`aspect_ratio` Target aspect ratio for the outpainted video. Choose a different ratio than the input video to extend the frame cleanly, e.g. use 16:9 for a 4:3 source.	`string`	no	`16:9`	`21:916:94:31:13:49:169:21`
`enable_prompt_expansion` Whether to enable prompt expansion.	`boolean`	no	`true`	—
`video_url` The source video to spatially outpaint.	`string`	yes	`—`	—
`video_write_mode` The write mode of the generated video.	`string`	no	`balanced`	`fastbalancedsmall`
`seed` Random seed for reproducibility. If None, a random seed is chosen.	`integer`	no	`—`	—
`video_quality` The quality preset of the generated video.	`string`	no	`high`	`lowmediumhighmaximum`
`frames_per_second` Frames per second of the generated video.	`number`	no	`24`	1–60
`sync_mode` If True, the media is returned as a data URI inline in the response. Useful for short-lived requests and tests.	`boolean`	no	`false`	—
`enable_safety_checker` Whether to enable the safety checker.	`boolean`	no	`true`	—
`num_frames` The number of frames to generate.	`integer`	no	`121`	9–481
`source_scale` Scale of the source video inside the target canvas before outpainting. 1.0 uses the largest source size that fits the target canvas, which is recommended for true aspect-ratio outpainting such as 4:3 to 16:9. Lower values create a zoom-out border around the source and are less stable.	`number`	no	`1`	0.25–1
`resolution` Legacy exact target canvas size. For new requests, prefer aspect_ratio with output_resolution; this field remains supported for backwards compatibility.	`ImageSize \| string`	no	`landscape_16_9`	—
`options` Platform-specific options for safety checks and Picsart Drive integration.	`object`	no	`—`	—
└`safety_checks` Safety check configuration.	`object`	no	`—`	—
└`enabled` Enable or disable text safety checks. Defaults to true.	`boolean`	no	`true`	—
└`drive` Picsart Drive storage options. When provided, the result will be saved to the user's Drive.	`object`	no	`—`	—
└`name` Filename to save the result as in Picsart Drive.	`string`	yes	`—`	—
└`attributes` Additional file metadata attributes.	`object`	no	`—`	—
└`folder` Destination folder in Picsart Drive.	`object`	no	`—`	—

Response

Over HTTP the output arrives inside a status envelope, at response.result. ai.apis.run() unwraps that envelope for you and resolves to { result, usage } instead. Either way the result itself is the vendor's own shape.

json

{
  "result": {
    "seed": 0,
    "video": {
      "url": "https://cdn.picsart.com/…/result.jpg",
      "content_type": "image/png",
      "file_size": 4404019,
      "file_name": "z9RV14K95DvU.png"
    },
    "prompt": "…"
  }
}

json

{
  "status": "success",
  "response": {
    "id": "wf_abc123",
    "status": "COMPLETED",
    "result": {
      "seed": 0,
      "video": {
        "url": "https://cdn.picsart.com/…/result.jpg",
        "content_type": "image/png",
        "file_size": 4404019,
        "file_name": "z9RV14K95DvU.png"
      },
      "prompt": "…"
    }
  }
}

LTXLTX 2.3 Outpaint API

Params

Result

Other models from LTX

Install the SDK

Authenticate

Make a request

Async (submit & poll)

Parameters

Response

Overview

Make a request

Async (submit & poll)

Parameters

Response

Rates