How to use Grok Imagine to create amazing images and videos, explains Elon Musk

Tech billionaire Elon Musk has drawn attention to xAI’s image tool, Grok Imagine, by sharing tips on how users can get better and more cinematic results. In a post on X on Sunday, Musk wrote, “How to use @Grok Imagine to create amazing images and videos,” while reposting a detailed guide by AI enthusiast @karatademada.

The post quickly gained traction, offering practical advice on how better prompting can significantly improve AI-generated images and videos.

Think like a director, not a typist

The guide says that the biggest mistake users make is writing very basic prompts. Instead of simply naming what appears in an image, users are advised to direct a scene, much like a filmmaker.

For example, instead of writing:
“A woman walking on a street.”

The post suggests using:
“Cinematic shot of a woman walking alone on a rainy Paris street at night, reflections of neon lights on wet pavement, filmed in 4K, directed by Christopher Nolan, atmospheric and moody.”

This works better because it explains the mood, setting and visual style, not just the action.

Use words that show emotion

According to the guide, Grok Imagine understands tone and emotion. Replacing plain words with more expressive ones can change the result completely.

Instead of:
“A happy girl under the sun.”

Users are encouraged to try:
“Close-up of a carefree young woman laughing under golden sunlight, wind blowing through her hair, summer energy, cinematic lens flare, warm tone.”

The advice is simple. Describe how the image should make the viewer feel.

Control the camera angle

The post also highlights the importance of camera and photography terms. These help Grok Imagine understand how to frame a scene.

Examples shared include:

“Wide establishing shot of a futuristic city skyline at dawn, soft mist, glowing reflections on glass towers, slow camera pan.”

“Low-angle cinematic shot of a hero standing on a rooftop overlooking the city, wind blowing coat dramatically, sun flares behind silhouette.”

“Tight close-up of a dancer’s face mid-performance, beads of sweat, emotional intensity, shallow depth of field.”

Such terms give the image a stronger sense of storytelling.

A simple structure for better results

To make prompting easier, the guide suggests thinking in five parts: what is happening, the visual style, the mood, the lighting and the camera view.

One example shared is:
“Scene: A samurai standing on a foggy mountain ridge.
Style: Cinematic realism inspired by Ridley Scott.
Mood: Stoic and powerful.
Lighting: Early dawn with soft mist.
Camera: Wide shot, 50mm lens, depth of field.”

Using this structure helps users get more consistent results.

Editing and expanding existing images

Grok Imagine can also improve images that have already been created. Users can ask the AI to add details or change the environment.

For example:
“Same image, but add gentle morning sunlight through the window, a cup of cappuccino on the table, and a Paris skyline reflection in the glass.”

Or a bigger change:
“Transform the same scene into a futuristic cyberpunk café with neon holographic menus and digital rain outside.”

This allows users to slowly build a story by improving the same image over time.

Improve through trial and error

The guide reminds users not to expect perfect images on the first attempt. Small changes can lead to much better results.

An example progression shared is:
“Portrait of a woman with flowers.”
“Portrait of a woman with yellow tulips under warm light.”
“Cinematic portrait of a woman holding yellow tulips, soft depth of field, 85mm lens, gentle morning glow.”

Each step adds more detail and clarity.