# Grok Imagine Video 1.5 for Product Videos: What xAI Actually Shipped

> For product marketers and founders evaluating AI video tools: what xAI's $0.02/sec Lite tier and pinned-frame controls do and don't document for product videos.

**Key takeaways**

- Grok Imagine Video 1.5 Lite charges $0.14 per second at 1080p, the same rate the full model charges at 720p.
- The full Grok Imagine Video 1.5 model supports up to 14 reference images, 3 voice references, and pinned first, last, and keyframe frames.
- xAI's documentation gives no spec for rendering small UI text or holding a steady camera on a screen.

Published Oct 21, 2026 · uscale · https://uscale.ai/blog/grok-imagine-video-1-5-product-videos

## The $0.14 line under the $0.02 headline

$0.14 a second. That is what Grok Imagine Video 1.5 Lite charges for 1080p on xAI's own model page. It is also exactly what the full Grok Imagine Video 1.5 model charges for 720p. The tier announced with a $0.02 headline costs the same as its bigger sibling once you ask for 1080p. And what comes back at that price is 720p output upscaled, not rendered natively.

So the low rate is real. Whether that makes Grok Imagine Video 1.5 for product videos a bargain, with a dashboard a buyer can actually read, is a different claim. The price list doesn't settle it.

[Video: Cinematic for Compute Exchange](https://uscale.ai/examples/compute-exchange-cinematic.mp4) — A live product filmed with camera moves, not generated from a prompt.
So what do xAI's own pages actually say about putting a product screen through this model?

## The short answer

uscale hasn't run Grok Imagine Video 1.5, so this section and the two after it read xAI's own documentation, not test results.

> **Generating hasn't replaced filming for product videos, because xAI documents new pricing and controls but not how the model renders small UI text or holds a camera steady on a screen.**

Start with the price, since that's where the headline is.

## What a second of Grok Imagine Video 1.5 costs at each resolution

Grok Imagine Video 1.5 Lite costs $0.02 a second at 480p, $0.03 at 720p and $0.14 at 1080p. The full model costs $0.08, $0.14 and $0.25 at the same three resolutions, according to xAI's own [Lite model page](https://docs.x.ai/developers/models/grok-imagine-video-1.5-lite) and [full model page](https://docs.x.ai/developers/models/grok-imagine-video-1.5), both dated October 2026.

![Grok Imagine Video 1.5: price per second by resolution: Lite 480p $0.02/sec, Lite 720p $0.03/sec, Lite 1080p (upscaled) $0.14/sec, Full 480p $0.08/sec, Full 720p $0.14/sec, Full 1080p (native) $0.25/sec](https://uscale.ai/blog-media/grok-imagine-video-1-5-product-videos/chart-grok-imagine-video-1-5-price-per-second--1200x520.svg "Sources: xAI docs — grok-imagine-video-1.5-lite model page; xAI docs — grok-imagine-video-1.5 model page")
The line worth remembering is in the middle of the chart. Lite at 1080p costs exactly what the full model charges at 720p. The reason is how each one gets there. Lite reaches 1080p by upscaling its 720p output, while the full model renders 1080p natively. At that price you are choosing between an upscaled 1080p frame and a native 720p one, not between a cheap tier and an expensive one.

That changes how to read the $0.02 headline, because it is the 480p rate. If your video will play at 720p or 1080p, it is not the rate you will pay. Start from the resolution your channel needs, then compare the two bars at that resolution. **A starting price tells you what the smallest clip costs. The rate at your delivery resolution tells you what your video costs.**

Price is only half of the tier choice. The controls a product brief relies on, reference images and pinned frames, are documented only for the full model, and the next section sets them out.

## The full model pins frames and takes references; Lite's docs list neither

The full Grok Imagine Video 1.5 model is the tier with controls. It takes up to 14 reference images per request, up from 7 on the older grok-imagine-video, plus up to 3 voice references. It pins the first frame, the last frame and up to 4 keyframes between them, on a 1/3-second grid strictly inside the clip. If you pair a starting image with `last_frame`, the model fills in the motion between the two. The classic model rejects `last_frame`.

Lite does text-to-video and image-to-video with lip-synced speech. xAI's docs don't list reference or pinning controls for it.

| Control | Video 1.5 | Video 1.5 Lite |
|---|---|---|
| Text- and image-to-video | Yes | Yes |
| Reference images | Up to 14 | Not listed |
| Voice references | Up to 3 | Not listed |
| Pinned first and last frame | Yes | Not listed |
| Mid-video keyframes | Up to 4, 1/3-second grid | Not listed |
| Clip length | 1 to 15 seconds | 1 to 15 seconds |
| Documented rate limit | Not listed | 10 requests/sec, us-east-1 and us-west-2 |

Pick the tier by the control your brief can't do without, then look at the price. A demo that must end on a specific screen needs a pinned last frame, so it belongs on the full model. A pinned frame fixes where a clip lands. It doesn't tell you whether the screen stays readable along the way, and that is where xAI's pages go quiet.

There's an obvious pushback, though, and it's about price.

## Isn't $0.02 a second cheap enough to just generate the product video?

The objection is fair. [Lite starts at $0.02 a second](https://docs.x.ai/developers/model-capabilities/imagine), the [lowest per-second rate in xAI's video lineup](https://docs.x.ai/developers/model-capabilities/video/overview). It handles text-to-video and image-to-video with lip-synced speech. At that rate a product marketer can generate many takes of a motion background or an abstract opening shot and discard most of them without straining the budget. For raw motion, cheap and disposable is exactly what you want.

That case supports the thesis. A low rate lowers the cost of trying, but it doesn't change what ends up in the clip. For a product video, xAI's own pages leave the important parts open. They describe no behavior for holding a camera steady on a screen. They give no spec for how small interface text renders. They mention no audio beyond lip-synced speech, so they say nothing about voice-over pacing or music under a walkthrough.

uscale hasn't run the model, so none of this is a measured failure. It is what the documentation leaves out. That gap is the useful finding. Cheap generation is a good reason to test Lite for plates and loops. It is not yet a reason to generate a dashboard instead of filming it.

So which of your videos go on which side of that line?

## How to decide this week which product videos to generate and which to film

The news changes one decision and leaves the other where it was. Here is how to split the work this week.

**Sort your planned videos into two piles.** The first pile is motion that doesn't have to match anything real: a background plate, a social loop, a few seconds of b-roll. The second pile shows your actual product: a dashboard, a signup flow, a pricing page. This is the quick win. A list in a doc is something you can finish within the hour. You're done when every video on your calendar sits in one pile.

**Test Lite on the motion pile only.** Both Grok Imagine Video 1.5 tiers cap clips at 15 seconds, so plan b-roll as short pieces. xAI's docs list Lite in two US regions for now, so check your access first. You're done when you've judged one clip against your brand.

**Film the product pile instead of generating it.** Nothing in xAI's documentation yet promises legible UI text or a held camera move on a live screen, and a launch or demo video depends on both. The free route is the screen recorder you already have and a script written in a doc. Another option is uscale. You paste your live site or a GitHub repo and approve a storyboard with every narration line. You get the real screens filmed, with voice, captions and music added. You're done when the script or storyboard is approved.

![The uscale start screen: paste your website and pick what the video should do](https://uscale.ai/blog-media/how-to/generate-form-1580x940.webp "Paste a live site or a GitHub repo; the real screens get filmed, not generated.")
If you do only one thing, move every product-screen video back onto a filming plan today rather than waiting for a model update.

Success looks like this:

- The launch page shows your real dashboard.
- Lite supplies motion only where nothing has to be accurate.
- Sales has one link to send.

For where this split applies, see [product explainer videos](https://uscale.ai/use-cases/product-explainer-video) and the [software products industry page](https://uscale.ai/industries/software-products).

**Try it:** [make a product explainer video](https://uscale.ai/use-cases/product-explainer-video) from your live site, with every narration line approved before filming.

## Before the docs catch up

1. **Watch xAI's video docs.** Before trusting Lite with anything UI-heavy, wait for a line on held camera moves or small on-screen text.
2. **Test Lite where text doesn't matter.** Background plates, social loops and b-roll are fair trials once you have access. The [earlier Lite breakdown](https://uscale.ai/blog/grok-imagine-video-1-5-lite-product-videos) covers that tier.
3. **Keep filming the real product.** A demo or launch video still needs the dashboard on screen as it actually looks.

For that third step, read how to [turn a URL into a product video](https://uscale.ai/blog/url-to-video), or watch [the example films](https://uscale.ai/examples) to see real screens filmed. Your first uscale video is free, and [the credit packs](https://uscale.ai/pricing) cover the ones after it. A cheaper clip is news, but the job is still a screen your buyer can read.

## Questions

### What is Grok Imagine Video 1.5?

Grok Imagine Video 1.5 is xAI's video generation model, which turns text or images into clips of 1 to 15 seconds, the kind of short clip a marketer might consider for a product demo or launch video. The full model renders 1080p natively, takes up to 14 reference images and 3 voice references, and pins first, last and mid-video frames. A cheaper Lite variant does text-to-video and image-to-video with lip-synced speech.

### What does the Grok Imagine Video 1.5 Lite tier cost?

Grok Imagine Video 1.5 Lite costs $0.02 per second of output at 480p, $0.03 at 720p and $0.14 at 1080p, according to xAI's model page. The 1080p rate matches what the full model charges for 720p, and Lite's 1080p is upscaled from 720p rather than rendered natively. For a video delivered at 720p or 1080p, the $0.02 headline rate is not the rate that applies.

### What new controls does the full Grok Imagine Video 1.5 model add?

The full Grok Imagine Video 1.5 model takes up to 14 reference images per request, up from 7 on the older grok-imagine-video, plus up to 3 voice references. It can pin the first frame, the last frame and up to 4 keyframes between them on a 1/3-second grid, and it fills in the motion between a starting image and a pinned last frame. xAI's docs list none of these controls for the Lite tier.

### Can Grok Imagine Video 1.5 make a product demo video?

Grok Imagine Video 1.5 is not yet documented as able to make a usable product demo video. It generates clips of up to 15 seconds and can pin a final frame, but xAI's pages give no spec for how small interface text renders or whether the camera holds steady on a live screen. It is a reasonable test for background plates, loops and b-roll; a demo of the real dashboard is still better filmed from the actual product.

### How much does a product video cost with uscale?

With uscale, the first product video is free: a new account gets 10 credits, or 20 with a work email, with no card needed. After that, credits are bought once with no subscription: Lite is $29 for 25 credits, about 2 product walkthroughs; Starter is $49 for 50 credits, about 5; Growth is $129 for 150 credits, about 15. uscale can also make the videos for you.

---
Source: https://uscale.ai/blog/grok-imagine-video-1-5-product-videos