NeuralBox

Grok Imagine 1.5: next-generation photo animation

Grok Imagine is the most popular video model among our users: tens of thousands of generations a month, most of them animating photos. Now the catalog has Grok Imagine 1.5 — the newest preview version of the xAI model, built for one scenario: the highest-quality animation from a photo.

One thing up front: 1.5 works from a photo only (image-to-video). For animating photos it costs about the same as the base version and adds 1080p. This post covers what changed, how much it costs and how to choose between the versions.

What Grok Imagine 1.5 is

Grok Imagine 1.5 is a preview of the newest generation of the xAI model. The scenario is the same as with the base version: upload a photo, describe the motion ("the person slowly smiles and tilts their head"), and get a video. The difference is quality: 1.5 delivers the most natural facial animation, and that's the reason to use it.

Text-to-video isn't available in 1.5 — the model works from photos only. For video from a text description, use the base Grok Imagine: it's still in the catalog at the same prices.

Image-to-video only
1.5 animates uploaded photos. Text-to-video is in the base version.
Best facial quality
The main difference from the base version is noticeably more natural facial animation.
6–15 seconds, up to 1080p
Choose a clip from 6 to 15 seconds at up to 1080p. On the website both versions offer 6–15 seconds; the base version tops out at 720p. Aspect ratios 16:9, 9:16, 1:1, 2:3 and 3:2.

Pricing and how it compares to the base version

There is no subscription: you top up your balance by card (via Stripe) or with crypto and pay per second of video. Prices are on the Basic plan ($5 = 150,000 tokens); on higher plans tokens are up to 40% cheaper.

At 480p and 720p, 1.5 costs only about 2% more than the base version, so for animating a photo there's little reason to stay on the old one. The base Grok Imagine remains the choice for text-to-video, which 1.5 doesn't do. Reliable mode for 1.5 (with a fallback to another provider if the main one fails) costs more: about $0.16/sec at 480p, $0.32 at 720p and $0.42 at 1080p.

ModeBase Grok ImagineGrok Imagine 1.5
480p, per second≈ $0.020≈ $0.020
720p, per second≈ $0.038≈ $0.038
1080p, per second—≈ $0.068
720p, 6-sec clip≈ $0.23≈ $0.23
Text-to-videoyesno — photo only

How to try it

Sign in with Telegram, Google or email, top up your balance, open the Video tab → Grok Imagine and select version 1.5. Upload a photo and describe only the motion — the model takes the appearance from the photo: "The people in the photo come to life: they blink, smile and hug each other."

A tip: describe only the motion and keep the prompt short — 1.5 takes everything else from the photo. Start at 480p or 720p to check the motion, then rerun at 1080p if you need the final clip in Full HD. A detailed overview of all modes and prompt examples is on the Grok Imagine on NeuralBox page.

Animate a photo in top quality

Grok Imagine 1.5 is already in the catalog: select version 1.5 on the Video tab.

Try Grok Imagine