MAI Image 2.5 AI Image Generator
One model for text-to-image and editing. Turn a written prompt into a finished image with strong prompt following, or add a reference image to rework that picture, made for characters, locations, and props described in plain language.
Key Features
Text-to-image generation with strong prompt following
Add one reference image to switch into edit mode and rework it
Reads plain-language descriptions of characters, places, and props
Cost scales gently with prompt length, so short prompts stay cheap
Works across photoreal, fantasy, and general illustration styles
Subscriber-only access on CharGen
Getting the Most from MAI Image 2.5
- Step 1
Describe the Whole Scene
MAI Image 2.5 follows written prompts closely, so spell out the subject, setting, mood, and lighting in one clear description.
- Step 2
Keep Prompts Focused
Cost rises a little as your prompt grows. Trim filler words and keep the description tight to control the gold cost per image.
- Step 3
Name the Style
Add a style cue such as 'oil painting' or 'cinematic photo' near the start of your prompt to steer the look.
- Step 4
Add a Reference to Edit
Upload one reference image and MAI Image 2.5 switches to edit mode, reworking that picture from your instructions while keeping the character and composition consistent.
- Step 5
Generate in Batches
Request several images at once to compare compositions before you commit to a single direction.
What You Can Make
Character Portraits
Turn a written backstory into a portrait for your party members, villains, and side characters.
Location Art
Visualize taverns, ruins, and city streets straight from a paragraph of description.
Prop and Item Concepts
Generate weapons, relics, and trinkets to hand out as loot or quest rewards.