Luma Ray 3.2 is now live in Picsart AI Playground

Luma Ray 3.2 is now live in Picsart AI Playground, and it lands with three distinct video modes in one place: text and image to video, video to video editing, and a generative reframe that resizes a clip while keeping its original audio. It is the latest version of Luma’s professional video model, built for precise, directable, production-ready output, and you can run all three modes today without a separate Luma subscription.

Meet Luma Ray 3.2

Luma Ray 3.2 is Luma AI’s most controllable video model to date. According to Luma’s official announcement, it generates clips up to 20 seconds long at 1080p and lets you place up to 16 keyframes inside a single clip, so you can choreograph exact camera paths and narrative beats instead of hoping a prompt lands. Luma describes it as “a precision tool, a highly directable, fully editable tool built for rigorous production environments.”

The model also tracks the full expressive state of up to eight faces at once, frame by frame, and supports native HDR generation with 16-bit EXR export so footage drops cleanly into existing post-production pipelines. Inside Picsart AI Playground, that professional control surface sits next to 100+ other models, so you can compare Luma Ray 3.2 against the rest on a single login and credit balance.

The 3 Luma Ray 3.2 modes in Picsart

Picsart exposes Luma Ray 3.2 as three separate modes, each with its own job. Here is what each one does and when to reach for it.

Text and image to video

The first mode turns a written prompt or a still image into a short video clip. Describe the scene you want, or upload a starting frame, and Luma Ray 3.2 animates it into motion. This is the classic generate-from-scratch workflow, useful when you have an idea or a single image and need movement around it. You can open the text and image to video mode here.

Video to video editing

The second mode edits footage you already have. Drop in a clip, describe the change in plain language, and the model returns a modified version, the same prompt-based editing approach that makes AI video editing fast for people who do not work in traditional editing software. It is built for restyling and altering existing shots rather than inventing new ones. The video to video editing mode lives here.

Reframe

The third mode is the one most creators will come back to. Reframe reshapes a shot after it was filmed: it changes the aspect ratio, extends the frame, and fills the new space with generative content, so a vertical clip can become a clean horizontal one without cropping the subject out. Luma notes that Reframe adapts the frame “while perfectly preserving the original lighting,” and in Picsart it also keeps the clip’s original audio intact. Try it in the Luma Ray 3.2 Reframe mode.

Why Reframe is the mode to try first

Reframe solves a problem every social creator runs into: footage shot for one platform almost never fits the next one. A vertical TikTok clip looks wrong on a widescreen YouTube banner, and a horizontal interview gets awkwardly cropped for Reels. Instead of re-shooting or hard-cropping and losing half the frame, Reframe rebuilds the missing edges with generative fill so the subject stays centered and the composition stays intact.

Two things make it genuinely useful in practice. First, it preserves the original lighting so the generated areas blend with the real footage rather than looking pasted on. Second, it keeps the source clip’s audio, which means a reframed video is ready to post without re-syncing sound. For anyone repurposing one video across vertical and horizontal placements, that combination turns a tedious manual job into a single generation.

Find Luma Ray 3.2 in Picsart AI Playground

All three modes live inside Picsart AI Playground, the unified workspace that brings video, image, and audio models together under one login and a single pay-per-generation credit balance. You can jump straight to text and image to video, video to video editing, or Reframe using the links above, then switch between models to compare results without leaving the page.

If your work is video-first, you can also reach Luma Ray 3.2 and the rest of the video lineup through the Picsart AI Video Generator. Either way, there is no separate Luma plan to manage: the model runs on the same Picsart credits as everything else in the Playground.

Start reframing with Luma Ray 3.2

Luma Ray 3.2 brings frame-level video control and a standout generative reframe into a workspace you already use. The fastest way to see what it does is to take one clip you have already shot, open Reframe, and turn it from vertical to horizontal in a single generation. Open Luma Ray 3.2 in Picsart AI Playground and start with a clip that needs a new aspect ratio.

Frequently asked questions

What is Luma Ray 3.2?

Luma Ray 3.2 is Luma AI’s latest professional video generation model. It produces clips up to 20 seconds at 1080p, supports up to 16 keyframes per clip for frame-level direction, and exports in HDR and 16-bit EXR for post-production pipelines. In Picsart it runs in three modes: text and image to video, video to video editing, and Reframe.

How many modes does Luma Ray 3.2 have in Picsart?

Three. Text and image to video generates clips from a prompt or still image, video to video editing modifies footage you already have from a text instruction, and Reframe changes a clip’s aspect ratio with generative fill while keeping the original audio.

Can Luma Ray 3.2 turn a vertical video into a horizontal one?

Yes. The Reframe mode changes the aspect ratio and fills the new space with generative content, so a vertical clip becomes a horizontal one without cropping the subject out, and it preserves both the original lighting and the original audio.

Do I need a separate Luma subscription to use Luma Ray 3.2?

No. Luma Ray 3.2 runs inside Picsart AI Playground on the same pay-per-generation credit balance as every other model, so there is no separate Luma plan to manage.