Updated October 22, 2026

Seedance 2.5

Seedance 2.5 is ByteDance's video generation model. Here is what it is, what it can do, and what we make with it.

What is Seedance 2.5?

Seedance 2.5 is a video generation model from ByteDance's Seed team. ByteDance describes it as a next-generation audio-video joint generation model, built for 30-second storytelling with precise reference control and powerful editing capabilities.

A single generation can run up to 30 seconds, and ByteDance says a video can be extended twice for longer stories. It describes the motion as smoother and more consistent than before, and the visuals as more realistic.

It also reads reference videos more closely. According to ByteDance it picks up the intention, framing and cinematic language of a reference clip and goes beyond copying the motion, and it handles editing requests more reliably than the earlier version.

What we use it for

Put Me In uses Seedance 2.5 to turn your photos into a short clip of a meme. For each meme we prepare a grey motion reference that holds the staging, the camera moves and the timing of the routine. Seedance 2.5 fills it with the people from your photos, so each of you plays your part in the shot.

Every clip is 15 seconds or shorter, and its file name and file metadata mark it as AI-generated. You don't write prompts and you don't need an API key: sign in with your email and password, or with Google, pick a meme, upload your photos, and the finished clip waits under My videos.

This is not a general text-to-video or image-to-video tool. We run set memes, and each one has been tuned for the shots and the pace of that routine. For the model itself, see ByteDance's own Seedance 2.5 page.

Seedance 2.5 prompts for meme videos

On Put Me In you don't write a prompt. If you prompt Seedance 2.5 or another image-to-video model yourself, here is how a meme scene is put together.

What goes into the prompt

Subject
Who is in the shot. Refer to each person by their photo (the person in photo 1) or by a name you use the same way every time.
Action
One main action per person. Long lists of moves tend to blur together.
Camera
How the camera moves, or that it stays locked off.
Setting and light
Where it happens and what the light is like.
Length and pace
Say the length and the shape first. A meme reads best at 15 seconds or less, with a few moves that each get time to land.

Clear, front-facing photos with the whole face visible work best, one person per photo. Leave out real names of celebrities, songs and brands, and use only photos of people who agreed.

Example prompts

  • Hotel Lobby

    A 15-second 9:16 vertical video in a bright orange studio with one microphone hanging from the ceiling. The person in photo 1 and the person in photo 2 stand side by side under the mic and perform a short, deadpan two-person routine: head turns in sync, small shoulder moves, taking turns to lean in to the mic. Locked-off camera, even studio light, minimal set.

  • Bad Times

    A 10-second 9:16 vertical video. The person in photo 1 stands in a sunny bedroom facing the camera and acts out an argument with it: frowns, points straight at the lens, throws up a hand in disbelief, then leans in close as if making a final point. Handheld phone framing, natural daylight, big comic expressions.

  • Ghostface

    A 5-second 9:16 vertical horror-film shot. In a dim bedroom at night, the person in photo 1 sits up in bed holding a phone to their ear, hears something, and slowly turns toward the open doorway, where a hooded figure in a pale mask stands in the dark. Cold moonlight, slow push-in, tense mood.

  • Smalltown Boy

    A 15-second 3:4 portrait video. The person in photo 1 and the person in photo 2 stand side by side in a living room lit in teal and pink and do an expressive hand-gesture dance in sync: hands framing the face, pointing, swaying, dramatic looks at the camera. Static camera, retro 1980s mood.

Here you skip all of this. Each meme has its own motion reference, tuned for its shots and pace, so you pick the meme, upload photos and get the clip.

Example prompts for every meme are on our MCP server for Claude, ChatGPT and Cursor. No key needed.

Memes made with Seedance 2.5 here

Each meme has its own prices for 480p, 720p and 1080p.

What it costs

From $6.99

Put Me In has no free plan, no trial and no subscription. You pay once per video. Hotel Lobby, the meme on sale now, costs $6.99 at 480p, $12.99 at 720p or $26.99 at 1080p, so a Seedance 2.5 video here starts at $6.99.

You pick a quality and see the exact price before you add your photos. The photos are checked first, and you pay once they pass. If a video can't be made, the payment is refunded in full.

Seedance 2.5 itself is ByteDance's model. ByteDance's Seedance 2.5 page links to the Dola app and to the BytePlus playground and API, and each of those sets its own terms.

See all prices

Seedance 2.5 questions

Seedance is a trademark of ByteDance. Put Me In is an independent site and is not affiliated with, endorsed by or sponsored by ByteDance.