Michelangelo
API DocumentationIntroducing Michelangelo, a pioneering tool that leverages the power of generative AI to create high-quality images, inspired by the renowned DALL-E 2. This cutting-edge technology is designed to generate images based on user prompts, requiring precise and highly descriptive inputs to deliver the best results. Michelangelo stands out by offering a unique negative prompt feature, which allows users to specify elements they don't want to include in the image. This is done by using affirmative wording, avoiding explicit negations, and giving special attention to things the user specifies. Another standout feature is its ability to adapt to different art styles, allowing users to either select from a list of styles or let the AI choose the best fit based on the prompt. If the choice isn’t clear, the tool defaults to a general art style, known for its appealing defaults. While the default resolution is set to 1024x1024 pixels, users have the option to specify the width and height, ranging from a minimum of 320 to a maximum of 1024. Michelangelo, in essence, offers a fresh take on image generation, bringing a blend of user control and AI innovation for an elevated creative experience.
Share:
Tags:
Example Prompts
Generate an image with the prompt "mountain sunrise" in a realistic art style with a width of
Create an image based on the prompt "futuristic cityscape" in a cyberpunk art style.
I want an image depicting "a knight battling a dragon" in a fantasy art style with dimensions
Can you make an image using the prompt "sakura tree in bloom" in an anime art style?
Generate an image with the prompt "steam-powered robot" in a steampunk style.
I'd like an image of "a spaceship in deep space" in a vector art style.
Please create an image using the prompt "underwater city" in a futuristic art style.
I want an image with the prompt "emoji party" in the emoji art style with a width of
Can you generate an image with the prompt "haunted forest" in a general art style?
I'd like an image with the prompt "ancient ruins in a desert" in a texture art style.
Description for AI
Generate images based on the user's prompt and provides the user with a link to the generated image. The prompt must be highly descriptive and as precise as possible to give the best results. It is possible to add a negative prompt which indicates to the AI what is NOT wanted for the image. This negative prompt must be written using affirmative wording, the model must never include explicit negations in it. For instance, if the image should not be ugly, the negative prompt for that would simply be "ugly". The model should give particular attention to things the user is not happy about and include them to the negative prompt paying respect to its due form. The negative prompt should not be provided otherwise. Different art styles are available to choose from. Depending on the prompt it is the model's responsability to figure out which art style fits best. The user should also be able to enforce the art style they want and ask for the list of available art styles. In case the art style to use is not obvious, the "general" art style should be preferred as it provides really good defaults. It is possible to explictly pass a width and height for the image. If these are not provided the default resolution is 1024x1024 pixels. The minimum value for the width and height is 320. The maximum is 1024. As a general rule, width and height should not be explicitly provided unless the user explicitly asks for a specific resolution.