Sign in

How to write better prompts for image generation?

Updated on October 2, 2024
Table Of Contents

To leverage the Image model’s capabilities fully, your prompts should be as detailed and as specific as possible.

Our model requires explicit, impactful trigger words. This precision ensures the model comprehends and executes the desired elements of your image accurately.

Key Trigger Words #

To optimize image quality and detail, consider incorporating these trigger words into your

prompts:

  • Skin Textures: Essential for detailed depictions of human or animal skin.
  • High Resolution: Ensures images are generated with utmost clarity. This token has three versions: “High Resolution”,”High-Resolution” or “High-Resolution Image.”
  • Cinematic: Perfect for creating dynamic, narrative-rich images with a film-like quality.

Components of a Prompt #

Craft each part of your prompt to be direct and descriptive, avoiding unnecessary verbosity:

Subject #

  • Definition: The primary focus of the image (e.g., person, animal, object).
  • Example: “Renaissance noblewoman”, “vampire hunter”.

Action #

  • Definition: Describes what the subject is doing, adding dynamism or narrative.
  • Example: “holding an ancient book”.

Environment/Setting #

  • Definition: The background or scene surrounding the subject.
  • Example: “in a dimly lit Gothic castle”.

Object #

  • Definition: Secondary items that enhance the subject or story.
  • Example: “wearing an intricate lace collar”.

Color #

  • Definition: Dominant colors or color schemes.
  • Example: “shades of deep red and gold”, “monochrome palette with stark contrasts”.

Style #

  • Definition: The artistic style or method of rendering.
  • Example: “in the style reminiscent of Vermeer’s lighting techniques”, “emulating a noir film”.

Mood/Atmosphere #

  • Definition: The emotional or atmospheric quality.
  • Example: “atmosphere of mystery”, “serene mood”.

Lighting #

  • Definition: Specific lighting conditions or effects.
  • Example: “bathed in soft, natural window light”, “dramatic shadows under a spotlight”.

Perspective/Viewpoint #

  • Definition: The angle or perspective from which the scene is viewed.
  • Example: “bird’s eye view”, “from a low angle”.

Texture/Material #

  • Definition: Prominent textures or materials visible in the image.
  • Example: “textures of rich velvet and rough stone”.

Time Period #

  • Definition: A specific era or historical period.
  • Example: “Victorian Era”, “futuristic 22nd century”.

Cultural Elements #

  • Definition: Elements that reflect specific cultures or traditions.
  • Example: “inspired by Norse mythology”, “traditional Japanese setting”.

Emotion #

  • Definition: The expressed emotion if the subject is sentient.
  • Example: “expression of deep contemplation”, “joyful demeanor”.

Medium #

  • Definition: Specifies the artistic medium or level of detail.
  • Example: “resembling a watercolor painting”, “crisp digital rendering”.

Clothing #

  • Guidance: Ensure prompts clearly specify subjects are clothed by describing the types and styles of garments, particularly when avoiding inappropriate content.
  • Example: “clothed in a Victorian gown”, “dressed in modern casual wear”.

Text #

  • Capabilities: Juggernaut has the ability to generate text, though accuracy decreases with sentence length and complexity. Use short, clear phrases enclosed in quotes to improve output.
  • Example: A Sign with the text “Peace”.

NSFW Content #

  • Avoidance: Explicitly include terms to exclude inappropriate content due to the model’s training data. Place sensitive terms in the negative prompts section.

Token Bleed and Multiple Subjects #

  • Challenges: When prompts combine multiple subjects or themes, there may be blending of elements, known as token bleed.
  • Solution: Be aware that multiple attempts might be needed to achieve desired separations in images depicting complex interactions like “Angel and Devil” and in not all cases possible but multiple seeds.

Examples of Prompts #

Here are some examples with the above trigger words and methodology incorporated for best end results.

PromptResult
Golden retriever in a lush green park with trees seen in the background, fur glistening under the soft afternoon sunlight, colors vivid green and golden, style realistic, mood serene, lighting natural and diffused, perspective close-up, texture soft fur and fresh grass, cultural elements joyful outdoor activity, natural bokeh effect
Orange cat enjoying her afternoon coffee, fur is soft and nicely combed, location Paris cafe, the Eiffel tower can be seen in the background, style realistic, lighting natural, perspective close-up, weather is sunny with no clouds in the sky, texture soft fur and wooden table, cultural elements calm afternoon cafe activity, natural bokeh background
Young woman reading a book in a cozy coffee shop, plants are positioned all around her making the atmosphere natural, brunette hair cascading over her shoulders, colors warm browns and soft whites, style candid photography, mood relaxed, lighting natural light streaming through a nearby window, perspective over-the-shoulder, texture soft wool sweater and glossy wooden table, a cup of coffee is placed near her on the table