Run the latest models all in one Sandbox 🏖️

Luma Ray 2 Text to Video

fal-ai/luma-dream-machine/ray-2
Ray2 is a large-scale video generative model capable of creating realistic visuals with natural, coherent motion.
Inference
Commercial use
Partner

Input

Additional Settings

Customize your input with more control.

Result

Idle

What would you like to do next?

Your request will cost $0.5 per 5 second.

Logs

Dream Machine Ray2 | [text-to-video]

Luma AI's Dream Machine Ray2 delivers realistic video generation with natural motion physics at $0.50 per 5-second clip. Trading rendering speed for motion coherence, Ray2 prioritizes physically accurate movement over raw generation time. Built for creators who need believable dynamics without manual animation rigging.

Use Cases: Product demos with realistic physics | Social media content with natural motion | Concept visualization for film pre-production


Performance

Ray2 positions as a quality-first text-to-video option, with scalable resolution pricing that charges 2x for 720p and 4x for 1080p output.

MetricResultContext
Base Cost$0.50 per 5s9-second clips cost 2x ($1.00)
Resolution Options540p / 720p / 1080p720p adds 2x cost multiplier, 1080p adds 4x
Aspect Ratios6 presets16:9, 9:16, 4:3, 3:4, 21:9, 9:21 native support
Loop ModeSeamless blendingEnd-to-start frame blending for continuous playback
Related EndpointsRay 2 Modify, Dream MachineVideo-to-video editing vs standard text-to-video variants

Motion Physics That Actually Look Right

Ray2 uses large-scale video training to understand natural motion patterns, how dust settles, how fabric moves, how objects interact with gravity. This differs from frame interpolation models that smooth between static images.

What this means for you:

  • Coherent motion across frames: Objects maintain consistent physics throughout the clip, not just visually similar appearance between keyframes

  • Natural lighting dynamics: Shadows and reflections update correctly as subjects move through space

  • Flexible duration control: Generate 5-second clips for rapid iteration, scale to 9 seconds when motion complexity demands it

  • Production-ready aspect ratios: Six native formats eliminate the need for post-crop workflows


Technical Specifications

SpecDetails
ArchitectureDream Machine
Input FormatsText prompts with detailed scene descriptions
Output FormatsMP4 video
Resolution Range540p to 1080p with proportional cost scaling
LicenseCommercial use via partnership

API Documentation | Quickstart Guide | Enterprise Pricing


How It Stacks Up

Luma Dream Machine (Text to Video) – Ray2 trades generation speed for enhanced motion physics at the same base cost structure. Dream Machine prioritizes faster turnaround for simpler motion scenarios where physical accuracy matters less than output volume.

MiniMax Video 01 Live – Ray2 focuses on realistic motion physics for natural scenes, while MiniMax emphasizes real-time generation speed for live production workflows. Both handle complex prompts, but optimize for different production constraints.

Luma Ray 2 Modify – The video-to-video variant extends Ray2's motion understanding to transformation workflows, letting you edit existing footage while maintaining the same physics coherence that defines the text-to-video model.