To make a Bad Times AI video, upload one clear photo to Bad Times Dance AI, pick a dance, choose a model, quality and format, and download the 13 or 15-second MP4 with the original song on it. There is no prompt to write and nothing to edit. Open the generator.
What are the four steps?
1. Add your photo
Use a front-facing photo in good light. The face, hair and outfit in the photo are the ones in the video, so choose the outfit on purpose. JPG, PNG or WebP up to 10 MB.
2. Pick the dance
Choose one of 6 dances: The Original, Solo Spotlight, Couple Duo, Friend Group, Dog or Cat, Pet Duo. Solo, Couple and Group run 13 seconds; The Original, Dog or Cat and Pet Duo run 15.
3. Set model, quality and format
The default is the Standard model (Seedance 2.0 Mini, made by ByteDance) at 720p in 9:16, which costs 10 credits. Higher tiers cost more credits and can keep the face closer to your photo. Formats are 9:16, 1:1 and 16:9.
4. Download and post
The MP4 comes with the original audio already added after rendering. Download it within 24 hours and post it.
Why not do it by hand?
You can, but it takes more than a prompt. Doing it yourself means finding a clean reference clip, holding an account with an image-to-video model, writing and tuning a long prompt, paying for several attempts and then putting the song back on the silent result and lining it up with the moves. Bad Times Dance AI does those parts once, per dance, and tests them.
| Bad Times Dance AI | Doing it yourself | |
|---|---|---|
| What you supply | One photo | Photo, reference clip, prompt, model account |
| Prompt writing | None, built in | Long, tuned by trial |
| Retries | Re-render if a result is off; a failed render returns its credits | Several paid attempts |
| The song | Original audio added for you | Find it, add it, sync it |
| Time | A few minutes | About 1 to 2 hours |
| Cost | From 10 credits ($6.99 for the smallest pack) | Model credits plus your time |
The prompt details are on the prompt page, and the full price list is on pricing. There is no free tier.
Which photo works best?
- Face the camera with eyes visible and no heavy beauty filter. The close-up at the end shows the face largest.
- Show the whole body for the wide first ten seconds of The Original. A tight face crop makes the model invent the rest.
- For couples, groups and two pets, stand side by side so nobody is hidden. Groups take three or four adults.
- For pets, use a sharp, front-facing photo with the whole animal in frame, as in the dog and cat guide.
- Only adults who agreed to the video, or your own pets. Photos of minors and public figures are not allowed.
How do I post it?
Export is vertical 9:16 by default, which fits TikTok, Instagram Reels and YouTube Shorts. The song is already on the file, but TikTok may still ask you to select the sound from its own library if you want the video on the sound's page. Label the post as AI-generated where the app offers that option. For where the format came from, read the trend explained; for why a one-tap app filter cannot do this, see the Bad Times AI filter.