Today we launched a cutting-edge image generation and editing model that can be used with Gemini apps, AI Studios, and Vertex AI. This update brings a huge advancement in character consistency. Accurate, conversational editing; and the ability to combine photography into whole new creations. To make the most of this update, here are some tips for creating more effective prompts for image generation and editing in Gemini.
Important features of Gemini image generation
Before you dive in, it’s helpful to get used to what’s improved in Gemini, so you can consider which use cases to try.
Consistent character design. Saves the appearance of a character or object across multiple generations and edits. Creative composition. Blend different elements, subjects and styles from multiple concepts into a single, unified image.local edit. Use Simple Language.design and appearance adaptation to accurately edit certain parts of the image. Apply style, texture, or design from one concept to another. Logic and reasoning. Use real-world understanding to generate complex scenes or predict the next step in the sequence.
Six factors to build effective prompts
You can get great results with Gemini from simple one or two sentences. However, to achieve the best results and unlock more subtle creative controls, consider including the following elements in the prompt:
Subject: Who or what is in the image? specifically. (For example, a stoic robot varistor with a sparkling blue optic; a fluffy calico cat wearing a small wizard hat). (Examples: extreme close-ups, wide shots, low angle shots, portraits). Action: What’s going on? (For example, Stride on your way through the field, brewing a cup of coffee, casting a magic spell). Location: Where is the scene going to take place? (For example, the future cafe on Mars, the messy alchemist library, the sun-drenched meadows of golden hour). Style: What is the overall aesthetic? (Examples: 3D animation, film noir, watercolor, photorealistic, product photos from the 1990s). Editing Steps: To modify an existing image, it’s directly and concrete (for example, change the man’s tie to green and remove the car in the background).
Prompt Example: A Showcase of Creative Techniques
Various prompt strategies allow you to unlock everything from light realistic editing to fantastical new worlds. Here are five techniques to try. Each of them contains important examples.
1. Save the character’s appearance.
Gemini can maintain the portrait of a person and character across a variety of poses, lighting and environments, and even apply the same character to new styles and surfaces. Here is an example of how to use a single character at multiple prompts in the same session.
Prompt 1: A whimsical illustration of a small, shining mushroom sprite. The sprite has a hat, wide, curious eyes, large bioluminescent mushroom cap for the body made of woven fabric.
