luma/ray-3.2

Luma's reasoning video model. Generates cinematic 5s or 10s video from text or images, with native HDR and EXR export for professional production pipelines.

29 runs

Readme

Luma Ray 3.2

Ray 3.2 is Luma AI’s reasoning video model. It generates short cinematic clips from a text prompt, a starting image, an ending image, or both — with native HDR and optional EXR export for professional production workflows.

What it does

  • Text-to-video — describe a scene and get a 5s or 10s clip.
  • Image-to-video — provide a start_image and/or end_image to anchor the generation.
  • Resolutions — 540p, 720p, or 1080p.
  • Aspect ratios — 9:16, 3:4, 1:1, 4:3, 16:9, 21:9.
  • HDR output — set hdr: true for true high dynamic range output (720p and 1080p only).
  • EXR export — set exr_export: true alongside hdr: true to get the ACES2065-1 EXR file used by professional color-grading workflows.
  • Looping — set loop: true for seamless looping 5s clips.

Constraints

  • 10s clips support text-to-video only — no start_image, end_image, or loop.
  • HDR is only available at 720p and 1080p, and only for 5s clips.
  • EXR export requires HDR.
  • Start and end images must be hosted at a public URL.

Pricing

Pricing is per video, by resolution × dynamic range × duration. See the model page for the full table.

Built by Luma AI

Ray 3.2 is part of Luma’s Ray 3 family, the first generative video model to produce true 10-, 12-, and 16-bit HDR in the ACES2065-1 EXR standard used in professional film and advertising pipelines. Learn more at lumalabs.ai.

Model created
Model updated