Start from a proven structure
Use a prompt as your first draft, then swap the subject, scene, lighting, camera angle, mood, or aspect ratio while keeping the structure that already reads clearly to the model.
Entdecke kuratierte Grok-Imagine-Prompts für Bilder und Videos, sieh dir Beispielergebnisse an und übernimm passende Prompts direkt in den Generator.
A massive epic cinematic battle in the sky above a burning futuristic city on Earth, viewed from low angle with dramatic perspective. A colossal white-and-pink high-tech mecha robot named "Xeno Leviathan Terraformer" dominates the scene, towering hundreds of meters tall, heavily armored with glowing purple and cyan accents, chest core blazing with intense light. It is firing a powerful bright blue energy beam from its chest and dual orange-yellow plasma beams from its arms downward. In front of the mecha, a fierce dark Godzilla-like kaiju with glowing red-orange dorsal spines and fiery lava effects on its body roars upward, and to the right a majestic golden dragon with large membranous wings, glowing eyes and fiery breath, also roaring aggressively. Both creatures are engulfed in smoke, fire, and explosions as they fight the giant mecha. Dramatic volumetric lighting, god rays piercing through thick smoke and clouds, lens flares, flying debris, burning city below with destroyed skyscrapers, visible Earth curvature and stars in the dark space background. Cinematic color grading, intense contrast, highly detailed, ultra-realistic textures, dynamic action pose, epic scale, masterpiece, best quality, 8k, photorealistic yet stylized anime/mecha aesthetic --stylize 750 --v 6
A black sports car racing through a narrow tunnel, engine echoing, headlights cutting through darkness. Swerves between vehicles and barely avoids crashing into barriers. Underground tunnel with flickering lights and tight walls. Forward tracking shot directly in front of the car, lights streaking past in rhythmic intervals, claustrophobic framing, intense speed perception.
Ultra detailed oil painting portrait of a woman standing in golden sunset light, her dress shimmering with Rococo floral textures and Art Deco symmetry, thick palette knife strokes forming glowing sky, warm tones dominating scene, hair flowing softly, emotional expression, cinematic composition, depth of field, museum quality, 8K.
A highly detailed 1/6 scale cyberpunk samurai robot action figure standing on a black hexagonal display base that slowly rotates 360 degrees on a transparent turntable. The figure is a menacing hooded cyborg ninja with a black tattered shredded cloak, exposed mechanical armor in dark gray and black, tactical pouches, straps, and weathered details. Its head is a sleek black robotic skull mask with intensely glowing bright red LED eyes and mouth that pulse with light. It holds a long katana with a glowing hot pink neon blade in one hand. The camera smoothly orbits around the figure at eye level, then moves in for dramatic close-ups: first the glowing red demonic face under the hood, then the mechanical arms and robotic hands gripping the sword, then the detailed legs and boots, and finally the back view showing the shredded cape and full armor. Professional studio lighting with soft white background, subtle reflections on the glossy floor, cinematic product photography style, ultra-realistic textures, sharp details, 4K, smooth 14-second rotating showcase video.
A busy red haired woman occasionally picking up her phone laughing and around her a green background begins to pixel eyes with numbers 13 13, 1313 expanding our mind and being able to predict what everyone else is doing to try to predict her
Create a dynamic live-action video featuring a person in the frame. The person remains relatively still with only natural subtle movements — slight head turns, natural blinks, breathing, small shifts in posture and gaze. No dancing, no hand waving
Rembrandt lighting combined with dark night hazy light creates mottled, diffused, blurred, and abstract light and shadow. Perfect color matching, high contrast, dark/hazy tone, strong film grain, high particle noise, layered blurred smudging, focus blur, motion blur, overlapping of figures and scenes, strong perspective, unique perspective, large areas of negative space (white space), minimalist background, and the screen incorporates local creative English information embellishments. 25
Grok Imagine prompt examples help creators move faster when they need working image prompts, cinematic video prompts, and reusable visual directions. This library combines copy-ready prompt text with uploaded preview assets so users can understand what kind of framing, lighting, camera motion, subject detail, and style language tends to perform well before opening the generator.
These examples are curated prompt references for creative inspiration. Results will still vary depending on model updates, settings, and the exact wording you use.
Damit die Seite schnell bleibt, werden zunächst nur kleinere Gruppen von Prompts geladen. Beim Scrollen werden automatisch weitere nachgeladen.
Use a prompt as your first draft, then swap the subject, scene, lighting, camera angle, mood, or aspect ratio while keeping the structure that already reads clearly to the model.
Image prompts usually need stronger visual composition and texture detail, while video prompts benefit from motion language, camera movement, pacing, and continuity instructions.
Every prompt card includes a use-now action that sends the prompt back to the homepage generator so you can immediately test, edit, and iterate without copying by hand.
It is a prompt library for users who want ready-to-use examples for image generation and video generation with Grok Imagine.
Yes. You can preview a prompt, copy it, or click the use-now button to open the homepage generator with that prompt prefilled.
They solve different problems. Image prompts focus on composition and visual detail, while video prompts also need motion, timing, camera movement, and consistency guidance.
Not always. Prompt examples are meant to improve your starting point, but exact outputs can still change with model behavior, settings, and prompt edits.