① Choose Yumi's identity references
Teach your AI who she is. These are identity anchors - not mood boards. Select the official references you want in the package. Defaults are ready for a first run.
First time? Keep it simple.
Leave Step 2 defaults alone. Choose your references here, then Ship it. Want control? Open Director controls in Step 2.
Before you start: Open the AI tool you plan to use in another tab or on your device. Leave it ready for image attachments and pasted text. You will send your finished Yumi package there in Step 3.
Checked images travel with your final package. Your own reference images are optional extras and stay in your browser until you export them.
Official Yumi references
These 8 images are Yumi's official identity anchors. Critical refs establish the core; the full set gives your AI more angles to work with.
Your own Yumi references optional
Privacy: Your files stay in your browser only. Nothing is uploaded to JHVinylShop or any server.
Add prior Yumi renders or extra angles in addition to the official refs above. This Builder always exports as Yumi (DAS-ACT-0001). Your extras are not validated against DAS rules and may cause consistency drift.
No custom images added yet. When selected, they join official refs in your Step 3 package.
② Shoot your shot
You call the shot. Yumi stays Yumi. Easy path: leave defaults and go to Ship it. Director path: open the controls below.
Easy path
Defaults are already configured for Yumi. Scroll to the bottom and continue to Ship it - or open Director controls only when you want to change the shot.
Director controls (pose, scene, camera, wardrobe, Always Yumi)
Always Yumi
Core identity - face, freckles, and beauty marks stay in place. Same person in every shot.
③ Ship it
Ship to AI: assemble your selected identity references, Yumi Rule Set, shot instructions, and structured data into one ready package. Download images first, attach them, then paste the text.
Send Yumi to your AI
- Confirm your AI tool is open and ready for attachments + paste.
- Download the complete package or contact sheet.
- Attach the images to your open AI.
- Paste the text package into the same AI session.
- Generate. Then use the built-in checklist to review the result.
The zip carries your selected refs plus any replacement refs and props. The text package tells your AI what to do with them.
Recommended order: download → attach images → copy text → paste → generate. Package image count updates as you change your selections.
Preview what will be copied or saved
Advanced (optional)
Limited attachments
Critical first: Portrait, Front, Detail, Expressions - then add Back, Left, Right, Full body.