Luma Modify Video in 2026: What Ray 3.2 Video-to-Video Does, What It Costs, and How to Run It

Luma Modify Video, now labeled Video To Video in Luma’s consumer app, is the video-to-video editing capability of the Ray 3.2 model, designed to restyle existing footage from text prompts while retaining varying amounts of its original motion, framing, and performance. Its nine strength presets are grouped into Adhere, Flex, and Reimagine bands, ranging from close structural preservation for retexturing to broad prompt-driven reinterpretation; flex_2 is the default setting. Ray 3.2 supports 5- or 10-second outputs at 540p, 720p, or 1080p, inherits the source aspect ratio, and allows up to 64 guide images or keyframes, while Luma’s API generally limits uploaded source clips to 18 seconds and 200 MB. Luma has shifted from earlier Ray 2 models and pixel-based pricing to Ray 3.2 and per-clip pricing, while Atlas Cloud offers the same model through a separate pay-per-run endpoint alongside competing editors such as Seedance, Kling, and Wan. Results can vary between runs, especially for face and character preservation at stronger settings, and Atlas Cloud outputs silent video, requiring users to restore audio separately.

No tracked trend matches for this post yet.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.