Positive Prompts

Positive Prompts

The positive prompt is the main input - it tells the AI what to generate. Each tag adds an element to the image, and their order determines priority.

Structure

Follow this recommended order for consistent results:

Example

by Chunie, solo, anthro, male wolf, grey fur, muscular, sitting, forest background, explicit

This prompt tells the AI: draw in Chunie's style, a single anthro male wolf with grey fur and a muscular build, sitting in a forest, explicit rating.

Tag Autocomplete

As you type in the positive (or negative) prompt box, a dropdown suggests matching e621 tags. Use the up / down arrows to move through the suggestions and Enter or a click to insert one. It helps you use real tags the models actually recognize, instead of guessing or inventing them.

• Start typing a few letters (e.g. mus) to see matches like muscular.
• Each suggestion shows its tag category color - the same categories listed in the Tag Reference.
• It works in both the positive and the negative prompt fields.

Quality Tags

Tags like best quality, masterpiece, details, HDR, 4K are mostly unnecessary. The current models already aim for high quality on their own, so adding them changes little. Just like the video guide, you can safely leave them out and spend those words on what actually matters: the character, pose and setting. (The older Yiff AI Basic model is the one case where a couple of quality tags can still help.)

Artist Tags

Artist tags have a big impact on the visual style. Here are some well-known artists that work reliably:

If an artist style feels too weak, increase emphasis: (by artistname:1.3)

Species Tags

Common species categories that work well: