---
title: "Grok Imagine: xAI's Image and Video Model Explained"
description: "Grok Imagine is xAI's model for both images and video. Get the modes, durations, costs and prompt tips, then run Grok Imagine on the FLORA canvas ✓ Try it free"
url: "https://flora.ai/academy/grok-imagine"
---

## [FLORA Academy](https://flora.ai/academy.md)

Master FLORA. Go beyond the fundamentals to the tools, techniques, models, and workflows behind professional creative work.

# Grok Imagine: xAI's Image and Video Model

Grok Imagine is xAI's generative model for both images and video. It runs on the FLORA canvas across six modes, from a 20 second image to a 15 second clip, alongside every other model you use.

**Want to try it?** Grok Imagine runs on the FLORA canvas in both image and video modes, so you can compare it against Nano Banana or Seedance on the same prompt before picking one. [Start free on FLORA](https://app.flora.ai/sign-up).

## What is Grok Imagine?

Grok Imagine is the image and video generation model built by xAI, the company behind Grok. It takes a text prompt, a reference image, or an existing video, and returns either a still or a clip. Most models do one or the other. Grok Imagine does both from the same prompt vocabulary, which is why it tends to get used for fast concept work where you are not yet sure whether an idea wants to move.

On FLORA it is available in six modes: text to image, image to image, text to video, image to video, video to video for editing an existing clip, and a references mode that builds video from frames you supply. Image runs finish in about 20 seconds and cost a few cents. Video runs take around 85 to 120 seconds and cost about 1.30 US dollars.

## Quick facts

- **Modes:** text to image, image to image, text to video, image to video, video to video, frames to video.
- **Image speed:** around 20 seconds. Quality variant around 20 to 30 seconds.
- **Video speed:** around 85 seconds text to video, 115 seconds image to video, 120 seconds for an edit.
- **Video duration:** 1 to 15 seconds, 6 seconds by default.
- **Resolution:** 480p or 720p, 720p by default.
- **Aspect ratios:** 16:9, 9:16, 1:1.
- **Cost:** about 0.04 US dollars an image, about 1.30 US dollars a video clip.
- **Provider:** xAI. Proprietary, so there is no self hosted option.

## What it is great for

- Concepting where you want a still and a moving version of the same idea without switching models.
- Short social cuts, since 15 seconds in one run covers most vertical formats.
- Editing an existing clip through the video to video mode rather than regenerating it.
- Cheap volume on the image side, where a run costs about four cents.

## Grok Imagine variants

FLORA carries a standard and a Quality build on the image side. Standard is about four cents a run and Quality about six, which is a small enough gap that most teams use Quality for anything that will be seen and standard for thumbnails and comps. On video there is a single build plus a References variant that takes frames as input, and an Edit build for changing a clip you already have.

## Example outputs

Text to image, standard build, about 4 cents

Text to image, Quality build, about 6 cents

Text to video, 6 seconds at 720p, about 1.30 dollars

### Copy and paste prompts

`Editorial product still on seamless paper, soft key light from camera left, crisp shadow, 50mm`

`Slow push in on the same product, 6 seconds, soft studio light, no camera shake, 16:9`

`Edit: regrade the clip warmer and add a slow drift to the right, keep the subject sharp`

## Parameters

| Name | Type | Default | Notes |
| --- | --- | --- | --- |
| prompt | text | required | Natural language description |
| duration | select | 6s | 1 to 15 seconds, video modes only |
| resolution | select | 720p | 480p or 720p |
| aspect_ratio | select | 16:9 | 16:9, 9:16 or 1:1 |
| image_url | upload | none | Required for image to video and image to image |

## Modes

| Mode | Estimated time | Cost | Typical use |
| --- | --- | --- | --- |
| Text to image | ~20s | ~$0.04 | Stills from a description |
| Image to image | ~20s | ~$0.04 | Restyling or editing a still |
| Text to video | ~85s | ~$1.30 | Short clips from a description |
| Image to video | ~115s | ~$1.30 | Animating a still you already have |
| Video to video | ~120s | ~$1.31 | Editing or regrading an existing clip |
| Frames to video | ~85s | ~$1.30 | Building a clip from supplied frames |

## Prompt tips

- Write the still and the clip from the same prompt, then add one motion clause for the video. Grok Imagine holds the look well across both.
- Name the duration you want. At 15 seconds, give the shot one clear action rather than several.
- Use 480p while you are exploring and switch to 720p for the version you keep. The cost is the same but the run is quicker.
- For an edit, describe only what changes. Everything you do not mention tends to stay put.

## Running Grok Imagine in FLORA

On the FLORA canvas Grok Imagine sits as a block, so a prompt or a reference feeds it directly and the output flows onward without leaving the board. Because it covers stills and video from one prompt, it fits the point in a workflow where you are deciding whether an idea should move.

Both halves have a matching technique. Use [text to video AI](https://flora.ai/techniques/text-to-video-ai.md) for prompt driven clips, or [image to video AI](https://flora.ai/techniques/image-to-video-ai.md) and the [AI image to video generator](https://flora.ai/techniques/ai-image-to-video-generator.md) to animate a still. On the image side, cut a subject out with the [background remover](https://flora.ai/techniques/background-remover.md) and finish on the [image upscaler](https://flora.ai/techniques/image-upscaler.md).

Models are interchangeable on the canvas, so run one prompt through Grok Imagine and a rival and compare. Try it against [Nano Banana](https://flora.ai/academy/nano-banana.md) and [Flux](https://flora.ai/academy/flux.md) on images, or [Seedance](https://flora.ai/academy/seedance.md) and [Kling](https://flora.ai/academy/kling.md) on video. Our rankings of the [best AI image generators](https://flora.ai/articles/best-ai-image-generators.md) and the [best AI video generators](https://flora.ai/articles/best-ai-video-generators.md) cover the wider field.

## Grok Imagine FAQs

### Who makes Grok Imagine?

xAI, the company behind Grok. It is a proprietary model, so there is no open weights version and no self hosted option.

### Is Grok Imagine free?

You can try it on FLORA's free tier before committing budget. Beyond that it draws from workspace credits. An image run is about four cents and a video clip about 1.30 US dollars, so cost depends heavily on which half you use.

### Can Grok Imagine make video as well as images?

Yes, and that is its main distinction. The same model covers text to image, image to image, text to video, image to video and video editing, so one prompt vocabulary works across stills and motion.

### How long can a Grok Imagine clip be?

Up to 15 seconds in a single run, with 6 seconds as the default. That is longer than most models manage in one pass.

### Is Grok Imagine good for commercial work?

The image side is cheap enough for volume and the video side holds a look across a short clip. As with any model, check the licensing terms that apply to your account and the rights on any reference you upload.