Runway Ruby Converts Generated Video Into HDR Masters
Runway’s new model converts uploaded or generated SDR footage into high-bit-depth HDR formats designed for professional post-production.
Generative video enters the finishing pipeline
Runway has introduced Ruby, a model that converts standard-dynamic-range video into high-dynamic-range output. It can process footage already uploaded to Runway as well as clips generated on the platform, and export up to 16-bit EXR image sequences or 10- and 12-bit ProRes and HEVC files.
The service supports the BT.2020 color space and professional transfer functions including PQ and HLG. These specifications matter because they preserve a substantially wider brightness and color range than ordinary web video, while EXR sequences provide the high bit depth and frame-level control expected in visual-effects and color-grading pipelines. Ruby is available to Max and Enterprise customers.
Ruby is not principally a new text-to-video generator. It addresses a later production stage: translating SDR source material into files that can move through professional editing, compositing and delivery systems. That distinction also shapes how its output should be assessed. A convincing HDR conversion must reconstruct highlight and color information that was not fully represented in the source, while avoiding unstable brightness, halos or color shifts from frame to frame.
Runway has not established through independent testing that converted footage contains the same recoverable scene information as material captured or rendered natively in HDR. The model’s exports should therefore be treated as generated interpretations of the SDR input, not as recovery of original sensor data.
Why it matters
Generative-video companies have largely competed on prompt adherence, motion and visual fidelity. Ruby moves the contest into post-production infrastructure, where compatibility with established formats can determine whether generated footage is useful beyond social media prototypes.
For studios, the appeal is a shorter path from an AI-generated clip to a grading or compositing suite, without first relying on a separate conversion service. More broadly, the release shows AIGC platforms expanding from asset generation toward end-to-end production workflows. Its lasting value will depend less on spectacular demonstrations than on temporal consistency, predictable color management and whether professional colorists can safely integrate the output into real projects.