---

name: luma-dream-machine-ray-2-video-reframe description: "Luma Dream Machine | Ray 2 | Video Reframe. Reframe videos to different aspect ratios using Luma Ray 2 AI. Triggers: video reframe, luma reframe, ray 2 reframe, change aspect ratio, reframe video, video crop ai, luma dream machine" allowed-tools: Bash(curl *), WebFetch

Luma Dream Machine | Ray 2 | Video Reframe

Reframe videos to different aspect ratios using Luma Dream Machine Ray 2. Intelligently reframes video content while preserving key subjects and composition, with optional prompt guidance for creative control.

Quick Start

Requires an each::labs API key. Get one at eachlabs.ai.

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "luma-dream-machine-ray-2-video-reframe",
    "version": "0.0.1",
    "input": {
      "video_url": "https://example.com/landscape-video.mp4",
      "aspect_ratio": "9:16"
    }
  }'

Parameters

Parameter	Type	Default	Description
aspect_ratio	string	9:16	The aspect ratio of the reframed video. Options: 1:1, 16:9, 9:16, 4:3, 3:4, 21:9, 9:21
grid_position_x	integer	-	X position of the grid for reframing
grid_position_y	integer	-	Y position of the grid for reframing
image_url	string	-	Optional URL of the first frame image for reframing
prompt	string	-	Optional prompt for reframing
video_url	string	-	URL of the input video to reframe
x_end	integer	-	End X coordinate for reframing
x_start	integer	-	Start X coordinate for reframing
y_end	integer	-	End Y coordinate for reframing
y_start	integer	-	Start Y coordinate for reframing

Examples

Reframe landscape video to vertical for social media:

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "luma-dream-machine-ray-2-video-reframe",
    "version": "0.0.1",
    "input": {
      "video_url": "https://example.com/travel-vlog.mp4",
      "aspect_ratio": "9:16",
      "prompt": "Keep the person centered in frame"
    }
  }'

Reframe to cinematic widescreen:

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "luma-dream-machine-ray-2-video-reframe",
    "version": "0.0.1",
    "input": {
      "video_url": "https://example.com/nature-scene.mp4",
      "aspect_ratio": "21:9"
    }
  }'

Reframe with custom coordinates:

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "luma-dream-machine-ray-2-video-reframe",
    "version": "0.0.1",
    "input": {
      "video_url": "https://example.com/group-video.mp4",
      "aspect_ratio": "1:1",
      "x_start": 100,
      "y_start": 50,
      "x_end": 500,
      "y_end": 450
    }
  }'

Related Models

• Luma Dream Machine | Ray 2 Flash | Video Reframe - Faster variant for quick reframes
• PixVerse v4.5 | Extend - Extend video duration
• Google Veo 3 - Text-to-video generation

Documentation

• each::labs Docs
• API Reference