Talk to Me Doja AI

How to make a Talk to Me Doja AI video

Upload one clear photo, pay once, and download a 12-second vertical video of you doing the Talk to Me Doja sway with the original sound. No sign-up, no filming.

Your video

$9.99 / video

No sign-up. Pay once with Stripe and your video starts automatically when you return. Save the private link below to find it again on any device.

Real output: 12 seconds, 9:16, made from one photo of an AI-generated fictional adult. Sound on for the original trend audio.

The steps

  1. Choose a photo of one adult with the face clearly visible. Waist-up or full body works best.
  2. Tick the consent box: it is your photo, or the adult in it agreed.
  3. Tap "Buy & generate" and pay with Stripe ($9.99 for one video). No account needed.
  4. You come straight back and generation starts by itself. It usually takes about 5 to 10 minutes.
  5. Download the MP4. The original trend sound is added on download.

Which photo works best

DoAvoid
One person, alone in the frameGroup photos
Face toward the camera, eyes visibleSunglasses, masks, hands over the face
Even daylight or soft indoor lightHeavy filters, beauty mode, dark rooms
Waist-up or full bodyTight face crops (the outfit gets invented)
At least 300 px on the short sideScreenshots of screenshots

What you get

SpecValue
Length12 seconds
FormatMP4, 9:16 vertical, 720p
SoundOriginal trend audio (sped-up "Like That")
WatermarkNone
MovesFollows the original TikTok: a relaxed start, then the cheek-cupping sway
ModelWan 3.0 in reference mode

Behind the scenes, the AI reads your photo for who appears (face, hair, outfit) and a real Talk to Me Doja TikTok for everything else: the moves, the camera, the timing and the anime stickers.

Posting it on TikTok

Caption and hashtags

talk to me Doja 🫶 #talktomedoja #speaktomedoja #likethat #dojacat #animedance #aivideo

  • Turn on "AI-generated content" in TikTok's post settings.
  • If the sound gets muted, add the trend sound from its TikTok sound page.
  • Post it as a reply to a friend's video for the "talk to me" joke to land.

FAQ

Can I make it for free?

No. Each render uses paid GPU time. Failed renders give your credit back automatically.

Can I put two people in one video?

Not yet. The current version replaces one person.

What if it does not look like me?

AI results vary. A sharper, front-facing photo usually helps. Clear technical defects can be reported to support.

Keep reading

By Kemi, independent developer. Last updated 2026-11-14