Kit
Same face, same room, same voice. Every post.
Build one believable AI person who looks, sounds and lives the same in every post.
More real output
- AI-generated performer
Talking to camera with a locked voice
AI-generated performerThe profiles-first character sheet 
A location grid, so the room stays the same - AI-generated performer
Product placement with your persona - AI-generated performer
The voice blueprint - AI-generated performer
Motion control: your performance, their body
How the video above was made
Same Mia, five places: car, golden hour, kitchen, gym, night · Five Grok Imagine clips, 1080p, 6-8 s, silent, cut to 9.5 s. Every start frame from GPT Image 2 with the same close-up and sheet. · Full example page
Five places, one identity: each start frame made in GPT Image 2 with Mia's close-up and sheet as references, animated silently in Grok Imagine, then cut to about 1.5 s each.
CAR (kept 2.5 s at 1.5x): 6 seconds, 9:16. Use the image as the opening frame. Handheld phone video, slight natural shake, one continuous take, no cuts, no talking. Face and hair unchanged from the image. No music, no text, no watermark. She squints into the sun, laughs, then pulls the sunglasses down onto her nose.
GOLDEN HOUR (1.8 s): 6 seconds, 9:16. Use the image as the opening frame. Handheld phone video, slight natural shake, one continuous take, no cuts, no talking. Face and hair unchanged from the image. No music, no text, no watermark. She keeps walking, glances back over her shoulder at the camera and smiles, her hair catching the low sun.
KITCHEN (1.6 s): 8 seconds, 9:16. Use the image (her in the kitchen holding the mug) as the opening frame. Handheld phone video, slight natural shake, one continuous take, no cuts, no talking. No music, no text, no watermark. She pushes the mug toward the lens until the logo fills the frame, then pulls it back, takes a sip and smiles. Mug shape, color and logo visible and unchanged. The hard morning sun beam stays as in the image.
GYM (1.6 s): 6 seconds, 9:16. Use the image as the opening frame. Handheld phone video, slight natural shake, one continuous take, no cuts, no talking. Face and hair unchanged from the image. No music, no text, no watermark. She catches her breath, wipes her forehead with the back of her wrist, then looks at herself in the mirror.
NIGHT, PHONE FLASH (1.9 s): 6 seconds, 9:16. Use the image as the opening frame. Handheld phone video, slight natural shake, one continuous take, no cuts, no talking. Face and hair unchanged from the image. No music, no text, no watermark. She tilts her head, gives a slow half smile into the flash, then looks away.What’s inside
- Building a believable persona from a candid base photo, not a studio portrait
- The profiles-first 3×3 character sheet, and 2×2 location grids for recurring places
- Eight preset content categories, and product placement with your persona
- The voice blueprint, writing for rhythm, and three ways to keep a voice consistent
- Video prompts for talking clips (Gemini Omni) and silent clips (Grok Imagine): talking to camera, new outfits, two characters and product demos
- Kling Motion Control: your performance on the character's body
- A failure-to-fix table, a one-page character bible and a ready-to-paste LLM character editor
Includes the interactive course + PDF
- 1Start here2 lessons
- 2Build the persona4 lessons
- 3Lock the identity9 lessons
- 4Make content in batches7 lessons
- 5Lock the voice5 lessons
- 6Put the character in motion11 lessons
- 7Reference4 lessons
42 lessons · about 46 min of reading
Refunds
If it isn’t useful, email support@aiofftheshelf.com within 14 days of buying for a full refund. Read the refund policy.
Questions
Which tools does it use?
GPT Image and Nano Banana Pro for images, Gemini Omni for talking clips and Grok Imagine for clips with no talking, Kling Motion Control for moving a real performance onto the character, and ElevenLabs for voice.
Can I build a character from a real person?
Only yourself, or someone who has given consent. An invented persona is the default. For talking-head videos of yourself, see the Clone Yourself Kit.
Does it cover AI disclosure?
Yes. The character bible states the persona is an AI character in the bio and captions, with an AI label on every post.
What do I get?
A 16-page PDF to keep. It's also included in The AI Video Playbook.