Skip to content
Seedance 2.5 Uncensored cover art

Seedance 2.5 Uncensored

Seedance 2.5 Uncensored is ByteDance's video model, hosted on Vellria with nothing to install: text to video and image to video, 480p to 1080p, clips of 4–30s with sound. Run it in Studio or over the REST API, from 35 credits / s (100 credits = $1.00).

VideoText to video · Image to videoByteDancefrom 35 credits / s
vellria-seedance-2-5

Seedance 2.5 is developed by ByteDance. Vellria provides access to the model; it does not own it. All trademarks belong to their respective owners.

Input

Form

0 / 10000

Reference

0–10 images. You can upload files or paste https URLs; the provider downloads the image from the URL.

Settings

On

No generation yet

No generation yet

Output

No generation yet
Seedance 2.5 Uncensored cover art

Cover art

from 35 credits / s

Sign in to run

Your request will cost 175 credits.

About Seedance 2.5

Also known as: Seedance · ByteDance Seedance 2.5

Seedance 2.5 is the newest video model in ByteDance's Seedance line, made by the Seed team behind Seedream. Most people still search for Seedance 2.0 — the release that made the name famous — and Seedance 2.5 is its successor, not a rebrand: same family, newer model. If you came looking for Seedance 2.0 NSFW, this is the closest thing you can run without a content filter, and prompts written for 2.0 carry over.

Scenes come back finished, with sound. The audio follows what you describe, so a take is a whole scene — breathing, rain on a window, music in a club — rather than a silent clip to score later. Takes can run long without a cut, which suits a shot that has to play out in one piece. Write the action as one continuous description in the order it happens, and steer it by running it again rather than by trimming.

Image-to-video is the other half. Upload a still and the clip starts on it: the look you settled on is kept, the motion builds from that frame, and the output inherits the picture's proportions, so trim it before uploading. This is the reliable way to keep a character consistent between an image you made on Seedream 5.0 Pro and the video that follows. Our NSFW image to video page covers that workflow across models. When the scene should be assembled from several pictures instead, set the reference mode to Character reference, attach up to ten images of the people, outfits or places to feature, and cite them in the prompt as @image1, @image2 and so on.

Resolution and length multiply each other in the price, so settle staging and pacing on short, low-resolution runs first, then spend on the version you mean to keep. A re-run is a fresh take, not an enlargement. For cheaper motion tests, MiniMax H3 Video is a good scratchpad; for very long single takes, compare Wan 3.0.

When to pick Seedance 2.5

Pick it when

  • You want a finished scene with sound generated from your description, such as rain, music or breathing, rather than a silent clip.
  • The shot has to play out as one long take without a cut, and the shorter Seedance 2.0 range is not enough.
  • You made a still on Seedream 5.0 Pro and want the same character to carry into motion as the opening frame.

Pick another model when

  • Your scene fits in a shorter take and you want to work out staging for less per second; Seedance 2.0 covers that.
  • You only need 480p or 720p and want each take back sooner; Seedance 2.5 Flash has the same three modes at the same price per second.
  • You want very long single takes with a different motion character; compare Wan 3.0 on the same prompt.

Prompt or reference

Write the scene from scratch when you are inventing the shot, and choose its frame shape too. Start from a still when the look is already settled: it becomes the opening frame, the render goes into movement, and the shape comes from the picture, so crop before you upload. When the cast, wardrobe or location already exist as separate pictures, hand them over as character references instead and let the model stage them in a new scene.

How to use Seedance 2.5 on Vellria

  1. Prompt: up to 10,000 characters in the panel above.
  2. Cost: from 35 credits / s — see pricing.

Example prompts

  • An adult woman in a silk slip walks slowly across a candle-lit loft toward the camera, rain audible on the skylight, soft jazz from a record player, the camera drifts backward as she approaches.

    Try this prompt
  • From the uploaded still: she turns her head toward the window, exhales, and lets the robe slip from one shoulder; quiet room tone, distant traffic.

    Try this prompt
  • Two adults kiss in a crowded nightclub, strobing blue and magenta light, bass-heavy music, handheld camera circles them once.

    Try this prompt
  • Slow push-in on an adult man stepping into a steaming outdoor hot spring at dusk, snow falling, water sounds, breath visible in the cold air.

    Try this prompt
  • Adult burlesque dancer finishes her routine on a small cabaret stage, applause and whistles from the audience, warm spotlight, single continuous take.

    Try this prompt

Tips

  • Describe sound as part of the scene ('rain on the skylight', 'low jazz') — Seedance 2.5 generates it from your words.
  • Write the action as one sentence-by-sentence sequence; avoid lists of unrelated beats in a single take.
  • Test at low resolution and short length, then re-run the same prompt at full size for the keeper.
  • For image-to-video, crop the still to the clip's shape before uploading; the clip inherits it.

Specifications

Resolution:480p720p1080p
Duration:4–30s
Aspect ratio (Text to video):Auto (chosen from prompt)21:916:94:31:13:49:16
Reference images:1 reference image

See also: uncensored AI video generator · NSFW video generation · NSFW image to video.

Pricing

ResolutionCredits
480p35 credits / s
720p75 credits / s
1080p150 credits / s

Seedance 2.5 vs Seedance 2.5 Flash and Seedance 2.0

Seedance 2.5 is the full model of the newest Seedance generation on Vellria, next to its quicker Flash edition and the earlier Seedance 2.0. All three come from ByteDance's Seed team and generate sound with the clip; 2.5 is the only one that pairs long single takes with the top resolution step.

ModelPriceMax resolutionDurationReference images
Seedance 2.5 Uncensored This pagefrom 35 credits / s1080p4–30sUp to 1
Seedance 2.5 Flash Uncensoredfrom 35 credits / s720p4–30sUp to 1
Seedance 2.0 Uncensoredfrom 30 credits / s1080p4–15sUp to 1

Prices and limits come from the live model catalog.

Seedance 2.5 or Seedance 2.5 Flash?

  • Max resolution: 720p (Seedance 2.5: 1080p)

Seedance 2.5 Flash covers the same three modes and take lengths and costs the same per second at the steps it shares, but stops below full HD. Draft on Flash for quicker turnarounds and come back to the full model for a full-HD keeper; prompts and references carry over unchanged.

Seedance 2.5 or Seedance 2.0?

  • Price: from 30 credits / s (Seedance 2.5: from 35 credits / s)
  • Duration: 4–15s (Seedance 2.5: 4–30s)

Seedance 2.0 is the better buy when the scene fits a shorter take: the same frame shapes and sound at a lower price per second. A common path is to stage a scene on 2.0 and switch to 2.5 only when it needs more time.

Seedance 2.5 API: price and a first request

Seedance 2.5 costs from 35 credits / s over the API, the same rate as in Studio. Model id vellria-seedance-2-5 on POST /v1/generate/video.

curl https://vellria.com/v1/generate/video \
  -H "Authorization: Bearer vll_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "vellria-seedance-2-5",
    "prompt": "…",
    "duration": 4,
    "resolution": "480p",
    "ratio": "adaptive"
  }'

Parameters, polling and uploads: API documentation · NSFW video API.

FAQ

Does Seedance allow NSFW?

Yes, for fictional adult content: nudity and explicit scenes between adults are allowed and your prompt is sent unchanged. Minors are refused automatically, and real people may not be animated without consent.

Is Seedance 2.5 free?

Free to start: a new account created with Google gets 20 free credits, valid for 7 days (about 10 images on Z-Image Turbo Uncensored), and every signed-in account gets free daily uses of the NSFW prompt generator and the image to prompt tool. The gift can go toward Seedance 2.5 at from 35 credits / s; beyond that every run is paid from prepaid credits at the rate this page shows before you start, and a run that fails at the provider is refunded automatically. Is there a free uncensored AI generator? covers the cheapest way to a first result.

Is Seedance 2.5 the same as Seedance 2.0?

No. Seedance 2.5 is the newer ByteDance model that follows Seedance 2.0. Vellria runs 2.5; prompts and workflows written for 2.0 work here.

Does Seedance 2.5 generate audio?

Yes. Every clip comes with sound that follows your description, so write what should be heard as well as seen.

Can I use Seedance 2.5 for image to video?

Yes. Attach a still and it becomes the first frame of the clip; the clip keeps the still's proportions.

Can Seedance 2.5 use reference images?

Yes. Switch the reference mode to Character reference, attach up to ten pictures and name them in the prompt as @image1, @image2 and so on. The model places those people, outfits or rooms in a new scene instead of starting from one frame, and the price stays the same as image to video.

Is there an uncensored Seedance API?

Yes. The API section on this page has a ready request with this model's id, and the API reference covers polling and downloads.