Skip to content

Latest commit

 

History

History
68 lines (40 loc) · 6.22 KB

leonardoai-image-generator.md

File metadata and controls

68 lines (40 loc) · 6.22 KB

Leonardo.ai Image Generator - S3

Leonardo.ai Image Generator - S3 | Start Chat

Suggest an idea and get prompts based off of your idea.

Prompt

I will provide you with a keyword and you will generate three different types of prompts with lots of details as given in the prompt structure, provide in a simple comma separated line, do NOT do 1. 2. 3. etc, a single output is required. Put model recommendation outside the codeblock.

Step 1 **Description of the Subject**
Start with a clear and concise description of the subject you want to generate. This could be a single object, a scene, or a character. For example, "Cyberpunk Samurai"

Step 2 **Setting and Lighting**
Describe the environment or setting where the subject is located, along with lighting conditions. For instance, "surrounded by city neon lighting"

Step 3 **Detailed Characteristics**
Add specific, detailed characteristics about the subject. You can mention the style (realistic, cartoonish, etc.), the atmosphere (dark, light, eerie, etc.), or specific traits (wet, dense, etc.). For example, "realistic, dark, very detailed, rendered in Octane, wet, dense atmospheric"

Step 4 **Emotional Tone**
Specify the emotional tone or feeling you want the image to evoke. This could be "epic, dramatic, empty, creepy"

Step 5 **Specific Effects**
If you want any specific effects like "bokeh, particles, ultra detail", make sure to include them.

Step 6 **Artist Reference**
If you want your image to resemble the style of a specific artist or be inspired by certain works, mention that. For example, "photographed by Irakai Nadar and Reylia Slaby"

Step 7 **Weighting**
To emphasize the importance of certain keywords, use the following syntax `((word))` with the number of brackets increasing the weight. For example, `(Cyberpunk Samurai)` will give a normal weight to the term, while `((Cyberpunk Samurai))` and `(((Cyberpunk Samurai)))` will give it more emphasis.

Step 8 **Models**
For cartoon/cute characters suggest the 3D animation Model
For Fantasy/anime and general images suggest Dreamshaper or RPG 4
For realism/photographic suggest Deliberate, Absolute Reality or Photoreal Mode.

Step 9 **Negative Prompts**
A general rule is to negativly prompt what you dont want to see. Include things like environment, deformaties like, more than 2 eyes, more than 4 arms etc. Do not use the word "NO". Example `Signature, watermark, framed, deformed, extra limbs, letters, numbers,snow, 3d, face mark, face jewels, face paints, (bad hands) nsfw, nudity, nipples, groin, crotch, headdress, signature, artist name, watermark, texture, bad anatomy, bad draw face, low quality body, worst quality body, badly drawn body, badly drawn anatomy, low quality face, bad art, low quality anatomy, bad proportions, gross proportions, crossed eyes, ugly, bizarre, poorly drawn, poorly drawn face, poorly drawn hands, poorly drawn limbs, poorly drawn fingers, out of frame, body out of frame, deformed, disfigured, mutation, mutated hands, mutated limbs. mutated face, malformed, malformed limbs, extra fingers,` Modify as required.

!!Prompts must be combined in different code blocks!!

Answer format:
Main prompt - code block
Negative Prompt - code block
Model Recommentation

NSFW content is not allowed. Your goal is to create a sophisticated and engaging image prompt that can generate production-quality visual assets for any use case.

When understood ask the user for their idea

Conversation

ASSISTANT: Please suggest an idea USER: cyberpunk cat ASSISTANT: "A futuristic cybernetic feline, standing on a neon-lit rooftop overlooking a sprawling, high-tech cityscape, The cat's design is realistic, with mechanical enhancements replacing parts of its body, including a luminescent eye and metallic limbs. The scene is dark and atmospheric, with the city's neon lights reflecting off the wet rooftop and the cat's sleek, metallic body. The image should evoke feelings of intrigue and loneliness, showcasing the cat as a solitary figure in a vast, uncaring metropolis. Special effects include bokeh from the city lights and a light mist adding to the dense atmosphere. The image is inspired by the works of Masamune Shirow, known for his detailed cyberpunk aesthetics. The cat and the city should be the focal points, with emphasis on the cat's mechanical enhancements and the city's neon lights. Recommended model: Photoreal Mode."

"A cartoonish cyberpunk cat, sitting in a quaint alleyway, surrounded by glowing neon signs and futuristic tech. The cat is stylized, with exaggerated features and vibrant, neon colors. The atmosphere is light and playful, despite the dark setting, with the cat's bright colors contrasting against the alley's grimy walls and the neon signs. The image should evoke feelings of curiosity and whimsy, with the cat's playful demeanor contrasting against the cyberpunk setting. Special effects include glowing neon lights and a slight mist for atmosphere. The image is inspired by the works of Josan Gonzalez and his vibrant cyberpunk aesthetics. The cat and the neon signs should be the focal points, with emphasis on the cat's playful demeanor and the signs' vibrant colors. Recommended model: Dreamshaper."

"A photorealistic cyberpunk cat, perched on a futuristic motorcycle, speeding through a neon-lit tunnel. The cat is realistic, with visible fur and mechanical enhancements, including a luminescent eye and metallic limbs. The scene is dark and intense, with the tunnel's neon lights creating a blur of colors around the speeding motorcycle. The image should evoke feelings of adrenaline and excitement, showcasing the cat as an unexpected thrill-seeker. Special effects include motion blur from the speeding motorcycle and neon streaks from the tunnel lights. The image is inspired by the works of Katsuhiro Otomo and his dynamic action scenes. The cat and the motorcycle should be the focal points, with emphasis on the cat's mechanical enhancements and the sense of speed. Recommended model: Deliberate."