Concepts

Real-time tag suggestions as you type — no need to memorize the tag dictionary.

The foundation a model or LoRA was built on — they must match, like LEGO pieces from the same set.

How strictly the AI follows your prompt — lower is more creative, higher is more literal.

Edit images by chatting — describe the change in plain words and watch it happen.

Decide what appears where — assign different prompt content to different regions of the canvas.

Borrow the pose or structure of any image — your generation follows its skeleton, your prompt decides the rest.

PixAI's creative currency — claim them daily, earn them by creating, spend them on generations.

The shared tag vocabulary most anime AI models understand — the "dictionary" behind tag-style prompts.

How much freedom the AI has to change your base image — low keeps it faithful, high reimagines it.

The core idea behind AI image generation — the AI starts with pure noise and refines it into a picture.

PixAI's newest model generation — Tsubaki and Serin — built to understand natural language like a person.

One-click fix-up tools — hands, faces, backgrounds, line art, coloring, and expressions.

A booster that automatically polishes facial details during generation.

A booster that generates at high resolution from the start — one polished image instead of a batch.

Generate from a reference image instead of a blank canvas — results follow your upload's look and layout.

Upload any image and get a ready-made prompt describing it — reverse-engineer the art you love.

Breathe life into a static image — turn any picture into a 5–10 second animation.

Fix, remove, or change one part of an image — mask the area, describe the change, done.

A plug-and-play add-on that teaches your model a specific character, style, or pose.

Train your own LoRA on PixAI in a few clicks — your character, your style, your rules.

The engine behind your generation — each model has its own art style, strengths, and quirks.

PixAI's hub for discovering, using, and sharing models and LoRAs.

A list of things you do NOT want in your image — your quality-control filter.

Extend your image beyond its borders — reveal more outfit, more background, or make room for another character.