Seedance 3.0 · AI Video Generator with Full Reference Control

Combine images, videos, audio, and text to produce cinematic videos with precise references, seamless extension, and natural language control.

Describe your video idea here...

Showcase

  • Multi-Modal Input
  • Reference Anything
  • 4-30 Seconds
  • Watermark-Free

Multi-Modal Input · Reference Anything · 4-30 Seconds · Watermark-Free

Core model

Seedance 3.0

Input types

Text · Image · Video · Audio

Best for

Marketing · Education · Storytelling

Seedance 3.0

Build multimodal scenes in one workspace. Bring images, video, and audio as references, then describe the subject, action, camera, and sound in the prompt and generate.

Black Friday

Save 50%
Price held for
30:00

50% off every annual plan until December 3 — applied automatically, no codes.

Black Friday

Save 50%

50% off every annual plan until December 3 — applied automatically, no codes.

Price held for
30
Min
:
00
Sec
Pricing

Choose a plan to buy credits

BASIC

Perfect for quick tests and first projects when you are just getting started

$29.80$14.90
/mo

Billed annually: $179 · Save $179 annually

$179/Year50% OFF

14,400 credits/Year

100 credits ≈ $1.24

14,400 credits/Year100 credits ≈ $1.24
  • 1,200 credits/month

  • Up to 60 videos/month

  • Up to 240 images/month

  • Seedance credit discount3.0 −23%Mini −33%

  • Nano Banana Pro

  • All AI Video Models

  • AI Music

  • Private Generation

  • Ad-Free

  • Priority Queue

  • Priority Support

  • Unlimited Storage

  • API access (credits shared with web)

  • Seedance 3.0

  • Commercial License

STANDARD

Best for professionals and frequent creators

$49.80$24.90
/mo

Billed annually: $299 · Save $299 annually

$299/Year50% OFF

30,000 credits/Year

100 credits ≈ $1.00 · 20% cheaper than Basic

30,000 credits/Year100 credits ≈ $1.00
  • 2,500 credits/month

  • Up to 125 videos/month

  • Up to 500 images/month

  • Seedance credit discount3.0 −23%Mini −33%

  • Nano Banana Pro

  • All AI Video Models

  • AI Music

  • Private Generation

  • Ad-Free

  • Priority Queue

  • Priority Support

  • Unlimited Storage

  • API access (credits shared with web)

  • Seedance 2.5

  • Seedance 3.0

  • Commercial License

🏆 Best Value

PRO

🏆 Best Value

The next step up for serious creators

$99.80$49.90
/mo

Billed annually: $599 · Save $599 annually

$599/Year50% OFF

72,000 credits/Year

100 credits ≈ $0.83 · 33% cheaper than Basic

72,000 credits/Year100 credits ≈ $0.83
  • 6,000 credits/month

  • Up to 300 videos/month

  • Up to 1,200 images/month

  • Seedance credit discount3.0 −23%Mini −33%

  • Nano Banana Pro

  • All AI Video Models

  • AI Music

  • Private Generation

  • Ad-Free

  • Priority Queue

  • Priority Support

  • Unlimited Storage

  • API access (credits shared with web)

  • Seedance 2.5

  • Seedance 3.0

  • Commercial License

MAX

For high-volume creators who need maximum credits

$199.80$99.90
/mo

Billed annually: $1,199 · Save $1,199 annually

$1,199/Year50% OFF

156,000 credits/Year

100 credits ≈ $0.77 · 38% cheaper than Basic

156,000 credits/Year100 credits ≈ $0.77
1x5x
  • 13,000 credits/month

  • Up to 650 videos/month

  • Up to 2,600 images/month

  • Seedance credit discount3.0 −23%Mini −33%

  • Nano Banana Pro

  • All AI Video Models

  • AI Music

  • Private Generation

  • Ad-Free

  • Priority Queue

  • Priority Support

  • Unlimited Storage

  • API access (credits shared with web)

  • Seedance 2.5

  • Seedance 3.0

  • Commercial License

Single Purchase • No Subscription

Credits are valid for 1 year from purchase. Buy anytime, use anytime.

Starter Pack

Quick top-up for small runs

$39.90
1,400 credits100 credits ≈ $2.85
  • Up to 70 videos

  • All features except commercial license

  • One-time purchase

  • No subscription required

  • Credits valid for 1 year

🔥 Most Popular

Creator Pack

Popular for regular usage

$79.90
3,200 credits100 credits ≈ $2.50
  • Up to 160 videos

  • Includes all features

  • One-time purchase

  • No subscription required

  • Credits valid for 1 year

Professional Pack

Best for larger batches

$199.90
10,000 credits100 credits ≈ $2.00
  • Up to 500 videos

  • Includes all features

  • One-time purchase

  • No subscription required

  • Credits valid for 1 year

Advanced Pack

Best for high-volume usage

$599
33,000 credits100 credits ≈ $1.82
  • Up to 1,650 videos

  • Includes all features

  • One-time purchase

  • No subscription required

  • Credits valid for 1 year

🏆 Best Value

Power Pack

Best for regular high-volume usage

$999
70,000 credits100 credits ≈ $1.43
  • Up to 3,500 videos

  • Includes all features

  • One-time purchase

  • No subscription required

  • Credits valid for 1 year

Max Pack

Best for high-volume usage

$2,499
180,000 credits100 credits ≈ $1.39
  • Up to 9,000 videos

  • Includes all features

  • One-time purchase

  • No subscription required

  • Credits valid for 1 year

Every credit pack

  • One-time purchase
  • No subscription required
  • Credits added instantly
  • Credits valid for 1 year

Popular for regular usage

Up to 160 videos*

Includes all features

* Video counts are estimates. Actual credits per video depend on the model, duration and resolution you choose.

Payment assistance: if you experience any issue during checkout, please contact our support team. support@seedances3.com

Pay safely and securely with

Payments are processed securely by Stripe. We never store your card details.

  • Mastercard
  • Visa
  • PayApple Pay
  • PayGoogle Pay
  • linkLink
  • JCB
  • American
    Express
    American Express
  • Discover
  • UnionPay

What Creators Are Planning For

Six working creators on what they get out of the open workspace now, and what the 30-second canvas changes for them.

“Thirty seconds in one pass is the number that matters to me. At fifteen I am always cutting the ending off; at thirty the reveal and the closing frame finally fit in the same take.”

Marcus Rodriguez

Marcus Rodriguez

Filmmaker

“Right now I burn most of a reference slot just describing the character. Splitting appearance, wardrobe and motion into separate sources is what I actually want from a deeper stack.”

Jessica Liu

Jessica Liu

Animation Director

“We stopped writing one enormous prompt and started pointing at things instead — this shot's lighting, that clip's camera move. The team argues less and ships more.”

Emily Watson

Emily Watson

Creative Director

“Thirty seconds in one pass is the number that matters to me. At fifteen I am always cutting the ending off; at thirty the reveal and the closing frame finally fit in the same take.”

Marcus Rodriguez

Marcus Rodriguez

Filmmaker

“Right now I burn most of a reference slot just describing the character. Splitting appearance, wardrobe and motion into separate sources is what I actually want from a deeper stack.”

Jessica Liu

Jessica Liu

Animation Director

“We stopped writing one enormous prompt and started pointing at things instead — this shot's lighting, that clip's camera move. The team argues less and ships more.”

Emily Watson

Emily Watson

Creative Director

“Thirty seconds in one pass is the number that matters to me. At fifteen I am always cutting the ending off; at thirty the reveal and the closing frame finally fit in the same take.”

Marcus Rodriguez

Marcus Rodriguez

Filmmaker

“Right now I burn most of a reference slot just describing the character. Splitting appearance, wardrobe and motion into separate sources is what I actually want from a deeper stack.”

Jessica Liu

Jessica Liu

Animation Director

“We stopped writing one enormous prompt and started pointing at things instead — this shot's lighting, that clip's camera move. The team argues less and ships more.”

Emily Watson

Emily Watson

Creative Director

“Thirty seconds in one pass is the number that matters to me. At fifteen I am always cutting the ending off; at thirty the reveal and the closing frame finally fit in the same take.”

Marcus Rodriguez

Marcus Rodriguez

Filmmaker

“Right now I burn most of a reference slot just describing the character. Splitting appearance, wardrobe and motion into separate sources is what I actually want from a deeper stack.”

Jessica Liu

Jessica Liu

Animation Director

“We stopped writing one enormous prompt and started pointing at things instead — this shot's lighting, that clip's camera move. The team argues less and ships more.”

Emily Watson

Emily Watson

Creative Director

“Changing one variable at a time is the whole discipline. Once I stopped rewriting the entire prompt between attempts I could actually tell what was improving.”

Mohammed Hassan

Mohammed Hassan

Digital Artist

“Letting the audio set the timing while images handle the performer and the location is the closest thing to how I plan a real shoot.”

Alex Turner

Alex Turner

Music Video Director

“More usable seconds per generation means more material to judge pacing against. Native 4K matters for the same reason — I need detail that survives the review, not just the preview.”

Olivia Martinez

Olivia Martinez

Video Editor

“Changing one variable at a time is the whole discipline. Once I stopped rewriting the entire prompt between attempts I could actually tell what was improving.”

Mohammed Hassan

Mohammed Hassan

Digital Artist

“Letting the audio set the timing while images handle the performer and the location is the closest thing to how I plan a real shoot.”

Alex Turner

Alex Turner

Music Video Director

“More usable seconds per generation means more material to judge pacing against. Native 4K matters for the same reason — I need detail that survives the review, not just the preview.”

Olivia Martinez

Olivia Martinez

Video Editor

“Changing one variable at a time is the whole discipline. Once I stopped rewriting the entire prompt between attempts I could actually tell what was improving.”

Mohammed Hassan

Mohammed Hassan

Digital Artist

“Letting the audio set the timing while images handle the performer and the location is the closest thing to how I plan a real shoot.”

Alex Turner

Alex Turner

Music Video Director

“More usable seconds per generation means more material to judge pacing against. Native 4K matters for the same reason — I need detail that survives the review, not just the preview.”

Olivia Martinez

Olivia Martinez

Video Editor

“Changing one variable at a time is the whole discipline. Once I stopped rewriting the entire prompt between attempts I could actually tell what was improving.”

Mohammed Hassan

Mohammed Hassan

Digital Artist

“Letting the audio set the timing while images handle the performer and the location is the closest thing to how I plan a real shoot.”

Alex Turner

Alex Turner

Music Video Director

“More usable seconds per generation means more material to judge pacing against. Native 4K matters for the same reason — I need detail that survives the review, not just the preview.”

Olivia Martinez

Olivia Martinez

Video Editor

What Is Seedance 3.0?

Seedance 3.0 is the multimodal generation model in the Seedance line. The idea it is built around is the one you can practise right here: give the model the images, video, and audio you want it to learn from, then use a text prompt for the subject, action, camera, mood, and sound — instead of forcing one long prompt to describe every visual detail.

Combine Text, Images, Video, and Audio

Use text to describe the result, images to define a person, product, or visual style, video to demonstrate movement and camera rhythm, and audio to guide speech, music, or atmosphere. Giving each input a clear role reduces guesswork and makes the first generation easier to direct.

Keep Characters, Products, and Style Consistent

Separate references can establish a character’s appearance, a product’s physical details, the environment, lighting, movement, and sound. This gives the model clearer visual boundaries and helps related shots feel like parts of the same project instead of unrelated clips.

Create a More Complete Video

When an idea needs more room, a single generation carries an opening, development, and ending across up to 30 seconds. Native 4K output covers the work where detail has to hold — products, lighting, typography, and close-ups — so you finish with a fuller video to keep editing rather than a clip to rebuild.

Seedance 3.0 vs Seedance 2.0

Seedance 2.0 already supports multimodal creation. Seedance 3.0 expands that foundation on four fronts: longer scenes, a much larger reference stack, stronger visual detail, and scene planning that no longer has to fit inside a short test clip.

Longer Single Generations: Up to 30 Seconds
2.0
4–15 sec3.0
4–30 sec

Seedance 2.0 is configured for 4–15-second generation. Seedance 3.0 extends a single generation to 30 seconds, leaving room for a setup, camera movement, primary action, transition, and ending beat. A longer canvas makes it possible to plan one coherent sequence instead of splitting the same idea across several short tests.

A Larger Multimodal Reference Stack
2.0
15 items3.0
50 items
30 images10 videos10 audio files

Seedance 2.0 supports up to 9 images, 3 videos, and 3 audio files. Seedance 3.0 raises this to 50 materials: up to 30 images, 10 videos, and 10 audio files. A stack that size lets character, product, environment, movement, and sound direction each occupy their own clearly defined source group.

Native 4K with Stronger Visual Detail
Native 4K
1080p

Same frame, 4× the pixels of 1080p

Native 4K output is there for the work that has to survive review: product texture, close-ups, launch media, lighting, and cinematic framing all hold up to inspection at a resolution where detail does not quietly disappear.

More Complete Scene Planning
2.0
Test clip
3.0
SetupActionTransitionEnding

The longer duration and larger reference stack make it easier to plan a complete scene instead of a short test clip. Seedance 3.0 also supports 3D spatial pre-visualization and camera blocking when subject placement, camera paths, and movement direction matter to the concept.

What Seedance 3.0 Does

The capabilities behind the workspace, and how each one changes the way a shot gets built.

A 30-Second Single Generation

One generation runs up to 30 seconds, with room for a setup, a main action, a transition, and a closing beat without stitching separate clips together.

Up to 50 Reference Materials

Bring up to 50 materials into one scene — 30 images, 10 videos, and 10 audio files — enough to give character, product, environment, motion, and sound each their own source group.

Native 4K Output

Render in native 4K for work where detail has to survive review: product texture, the printed text on packaging, fabric weave, and close-ups that dissolve at lower resolution.

Reference-Led Direction

Point at what you want carried over — a camera path, a performance, a lighting mood — instead of packing every visual detail into one long prompt.

Consistency Across Shots

Give a character or product its own reference group and repeat its defining details in the prompt, so related shots read as one project rather than a set of unrelated clips.

3D Spatial Pre-Visualization

Built for camera blocking and spatial previs, where subject placement and camera paths need to be argued out before anyone books a shoot day.

Where a Longer, Reference-Led Model Changes the Work

Labels like "social media" or "e-commerce" hide the differences that matter in production. These are the specific jobs a 30-second canvas and a deep reference stack are built to serve.

TikTok Effects and Vertical Hooks

Earn the first two seconds: a transformation, a surreal effect, a music-led cut. Images fix the subject, a motion reference carries the effect, audio sets the rhythm.

Vertical HooksTransformationsMusic-Led CutsTrend Testing

Marketplace Product Demonstrations

Build a product-led video from authorized packshots and material details. Show what the item is and how it is used, without effects that outshine the product.

Packshot FilmsUsage DemosMaterial DetailListing Assets

Character IP and Animated Shorts

Hold one character across scenes by giving appearance, wardrobe, and movement their own references, then explore styles without redrawing the character each time.

Character DesignAnimated ShortsStyle ExplorationSeries Continuity

Campaign Pre-Visualization

Test composition, camera paths, transitions, and the closing frame before a shoot is booked, so the discussion happens over footage rather than a deck.

Camera BlockingShot PlanningClient PrevisConcept Proofing

Music Performance and Rhythm Edits

Let an audio reference set the timing while image and video sources establish the performer, the location, and the camera language around them.

Beat TimingPerformance CutsStage LanguageAudio Sync

Training and Explainer Sequences

Turn a procedure into a watchable sequence where the steps stay in order and the subject stays recognizable from the first frame to the last.

Procedure WalkthroughsOnboardingSafety BriefingsCourse Assets

Property and Interior Walkthroughs

Move stills into motion so a room reads as a space: where the light falls, how the layout connects, what the finish actually looks like.

Room ToursLayout ReadsFinish DetailListing Video

Storyboard to Motion Test

Take panels that already exist and put them in sequence and in motion, to find out which beats hold and which need reworking before production.

Panel SequencingBeat TestingTiming ChecksPitch Reels

Brand Identity Films

Keep wardrobe, product, environment, and lighting direction in separate reference groups so a brand's visual rules survive from one film to the next.

Visual IdentityLaunch FilmsBrand RulesSeries Consistency

1 / 9

How to Create AI Videos with Seedance 3.0

1

Decide Where It Will Be Published

A marketplace product demo, a vertical hook, and a campaign previs need different aspect ratios, pacing, and detail levels. Settle this before writing a word of prompt.

2

Give Each Reference One Job

Let one source define the character, another the product, another the camera rhythm. Separated sources leave far less to guesswork than a single crowded image.

3

Write the Scene in Order

Subject, setting, action, camera, light. State when the camera moves and what the last frame should say, then change one variable at a time as you refine.

Frequently Asked Questions

How Seedance 3.0 compares to the previous model, what you can create today, and how credits and licensing work.

Free every day. Every signed-in account can create one Seedance 2.0 Mini video each day at no cost, with no payment method required. Seedance 2.0 and Seedance 2.5 use credits; see Pricing for current plans and credit packs.

  • Seedance 2.0: 4–15 second videos guided by text, image, video, and audio references.
  • Seedance 2.5: clips up to 30 seconds, native 4K output, and up to 50 reference files.
  • Seedance 2.0 Mini: a lower-cost option for quick drafts, with one free video every day.

Seedance 3.0 extends a single generation from 4–15 seconds to up to 30, expands the reference stack from 9 images, 3 videos, and 3 audio files to up to 50 materials, and adds native 4K output. The longer canvas makes full scene planning practical instead of compressing an idea into a short clip.

Open the workspace on this page, choose how you want to start — media, image, or text — then describe the subject, action, camera, and sound in your prompt. Uploaded references can be cited as @1, @2, @3 directly in the prompt.

The Seedance 3.0 workspace takes text prompts alongside image, video, and audio references in one place, so you can build multimodal scenes today.

Yes. Credits belong to your account rather than a single model, so any unused balance carries over and stays available as new models are added to the workspace.

The workspace shows the current limit for your account before you submit a generation. Use the live interface as the source of truth, since limits differ between models.

Videos generated on paid plans are watermark-free. Commercial usage rights for client work, advertising, and social media depend on the plan: they come with all annual plans, the monthly Standard, Pro and Max plans, all Team plans, and credit packs from the Creator Pack up. The free tier, the monthly Basic plan, the Starter Pack and the Trial Pack are for personal, non-commercial use only. You still need to own or license whatever source material you upload.

No. The daily free video applies to Seedance 2 Mini only. Every other video model, the image models, AI Music, Virtual Try-On and the upscalers use credits. Content created on the free tier is for personal, non-commercial use.

Yes. Commercial licenses come with all annual plans, the monthly Standard, Pro and Max plans, all Team plans, and credit packs from the Creator Pack up. Content created on the free tier, the monthly Basic plan, the Starter Pack or the Trial Pack is for personal, non-commercial use only. The Pricing page lists the details.

Premium plans include watermark-free videos for professional use. The Pricing page shows which plans apply to you.

Payments are processed securely through Stripe, and checkout shows the payment methods available for your region and purchase. Credits are shared across every model and the API. One-time credit packs stay valid for 1 year, and subscriptions can be paused at any time to avoid losing unused credits. If a generation fails for technical reasons, it is retried automatically and the credits are refunded in full when it cannot complete.

Have more questions? support@seedances3.com →

Stripe payments

DMCA/CCPA friendly

0+

Used by creators & shops

0+

Videos Generated

Start Creating with Seedance 3.0

Build your reference set, settle your prompt, and generate — everything you need is in the workspace above.