HiNow LogoHiNow

LTX-2.3 22B

Model Information

Display Name: LTX-2.3 22B

API Model ID: lightricks/ltx-23-22b

Category: Image To Video

Description: LTX-2.3 22B is Lightricks' latest 22 billion parameter video generation model, producing high-quality video from images with optional start and end frame interpolation. Features built-in audio synthesis, camera control LoRAs, and advanced spatiotemporal guidance. **Key Features:** - 22B parameter Diffusion Transformer architecture - Image-to-video with start and end frame interpolation - Built-in audio generation - Camera control LoRAs (dolly, jib, static) - Multi-scale generation for enhanced quality - Configurable frame count, FPS, and resolution - Multiple output formats (MP4, VP9, ProRes, GIF) **Capabilities:** - Image-to-video generation - Start + end frame interpolation - Automatic audio synthesis - Camera movement control - Negative prompt support - Prompt expansion for better results - Seed control for reproducibility **Best For:** - High-quality video generation from images - Video transitions between two keyframes - Creative video content with camera movements - Marketing and social media videos - Video prototyping with audio **Input Requirements:** - Image: Start frame (required) - End Image: End frame for interpolation (optional) - Prompt: Motion/scene description (required) **Parameters:** - aspect_ratio: 16:9 (default), 9:16, 1:1, 4:3, 3:4, 3:2, 2:3, 21:9 - num_frames: frame count (default 121) - fps: frames per second (default 24) - num_inference_steps: denoising steps (default 40) - camera_lora: dolly_in/out, dolly_left/right, jib_up/down, static, none - generate_audio: true/false (default true) - acceleration: none, regular, high, full - video_quality: low, medium, high, maximum **Pricing:** - Per video megapixel (width x height x frames) - Cost scales with resolution and frame count - Example: 1280x720, 121 frames ≈ /bin/bash.25 **Limitations:** - Higher resolutions and frame counts increase cost - Generation time varies with acceleration setting - Camera LoRA movements are preset, not custom paths

How to Use This Model

To use LTX-2.3 22B via the HInow.ai API, use the model ID: lightricks/ltx-23-22b

API Request Example (Image Generation)


POST https://api.hinow.ai/v1/images
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

{
  "model": "lightricks/ltx-23-22b",
  "prompt": "Your image description here"
}

Pricing

  • total: $0.25

Available Parameters

  • aspect_ratio: Output video aspect ratio (Options: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 21:9)
  • num_frames: Number of frames (more frames = longer video) (Options: 61, 81, 97, 121, 161)
  • fps: Frames per second (Options: 16, 24, 30)
  • num_inference_steps: Number of denoising steps (Options: 20, 25, 30, 35, 40)
  • generate_audio: Generate synchronized audio (Options: true, false)
  • negative_prompt: What to avoid in the video
  • seed: Random seed for reproducibility
  • image_strength: How strongly to follow the input image (Options: 0.3, 0.5, 0.7, 0.85, 1.0)
  • camera_lora: Camera movement preset (Options: none, pan_left, pan_right, tilt_up, tilt_down, zoom_in, zoom_out, orbit_left, orbit_right)

Quick Reference

To use this model, set: "model": "lightricks/ltx-23-22b"

Featured: No

Documentation: https://hinow.ai/models/lightricks/ltx-23-22b

API Endpoint: https://api.hinow.ai/v1

Back to Models

LTX-2.3 22B

lightricks/ltx-23-22b

$0.250
per image

About

LTX-2.3 22B is Lightricks' latest 22 billion parameter video generation model, producing high-quality video from images with optional start and end frame interpolation. Features built-in audio synthesis, camera control LoRAs, and advanced spatiotemporal guidance.

Key Features:

  • 22B parameter Diffusion Transformer architecture
  • Image-to-video with start and end frame interpolation
  • Built-in audio generation
  • Camera control LoRAs (dolly, jib, static)
  • Multi-scale generation for enhanced quality
  • Configurable frame count, FPS, and resolution
  • Multiple output formats (MP4, VP9, ProRes, GIF)

Capabilities:

  • Image-to-video generation
  • Start + end frame interpolation
  • Automatic audio synthesis
  • Camera movement control
  • Negative prompt support
  • Prompt expansion for better results
  • Seed control for reproducibility

Best For:

  • High-quality video generation from images
  • Video transitions between two keyframes
  • Creative video content with camera movements
  • Marketing and social media videos
  • Video prototyping with audio

Input Requirements:

  • Image: Start frame (required)
  • End Image: End frame for interpolation (optional)
  • Prompt: Motion/scene description (required)

Parameters:

  • aspect_ratio: 16:9 (default), 9:16, 1:1, 4:3, 3:4, 3:2, 2:3, 21:9
  • num_frames: frame count (default 121)
  • fps: frames per second (default 24)
  • num_inference_steps: denoising steps (default 40)
  • camera_lora: dolly_in/out, dolly_left/right, jib_up/down, static, none
  • generate_audio: true/false (default true)
  • acceleration: none, regular, high, full
  • video_quality: low, medium, high, maximum

Pricing:

  • Per video megapixel (width x height x frames)
  • Cost scales with resolution and frame count
  • Example: 1280x720, 121 frames ≈ /bin/bash.25

Limitations:

  • Higher resolutions and frame counts increase cost
  • Generation time varies with acceleration setting
  • Camera LoRA movements are preset, not custom paths

Capabilities

Image To Video

Parameters

aspect_ratio

Output video aspect ratio

1:116:99:164:33:43:22:321:9
num_frames

Number of frames (more frames = longer video)

618197121161
fps

Frames per second

162430
num_inference_steps

Number of denoising steps

2025303540
generate_audio

Generate synchronized audio

truefalse
negative_prompt

What to avoid in the video

seed

Random seed for reproducibility

image_strength

How strongly to follow the input image

0.30.50.70.851.0
camera_lora

Camera movement preset

nonepan_leftpan_righttilt_uptilt_downzoom_inzoom_outorbit_leftorbit_right

Code Examples

curl -X POST https://api.hinow.ai/v1/videos \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $HINOW_API_KEY" \
  -d '{
    "model": "lightricks/ltx-23-22b",
    "prompt": "A cat playing with a ball",
    "images": ["https://example.com/image.jpg"],
    "parameters": {
      "aspect_ratio": "1:1",
      "num_frames": "61",
      "fps": "16",
      "num_inference_steps": "20",
      "generate_audio": "true",
      "negative_prompt": "",
      "seed": "",
      "image_strength": "0.3",
      "camera_lora": "none"
    }
  }'