Official model reference

Seedance 2 on Veyloom

Use Seedance 2 workflows currently available on Veyloom for text-to-video, image-to-video, video-to-video, and frame transition creation, with reference clips analyzed separately from Veyloom output.

Official ByteDance media and competitor media are not embedded or rehosted. Reference clips below are attributed separately and are not Veyloom output.

Motion evidence

See the model in motion

Video analysis

How to read Seedance 2 motion quality

The left/right layout keeps the clip and the explanation together, so each example has a concrete reason to be on the page.

Veyloom Seedance Fast sample

Atmosphere, depth, and reference hold

Applicable scenario:

Use this when an image-to-video project starts from a still environment frame and needs rain, mist, particles, or soft ambient motion without losing the original composition.

Atmosphere, foreground motion, and source-image hold

  • The signal is whether atmosphere, depth, and foreground motion stay readable from the source image.
  • Separate environmental motion from camera movement so the still frame does not drift.
Veyloom Seedance Fast sampleAtmosphere, depth, and reference hold
Veyloom Seedance Fast sampleInterior camera move and spatial stability

Veyloom Seedance Fast sample

Interior camera move and spatial stability

Applicable scenario:

Use this when the source frame is an interior, showroom, or architectural scene. Control camera path, parallax, and light behavior while keeping the room layout stable.

Interior layout stability and controlled camera movement

  • Interior shots expose whether layout can stay stable while camera push or parallax is added.
  • Name the path, pace, and lighting behavior instead of asking for a generic cinematic result.

Reference clips for concrete motion analysis

Reference motion clip

Locked-camera seasonal transformation

Applicable scenario:

For landscape, city, or location videos, lock the frame first and then describe the allowed change: season, weather, time of day, traffic, or crowd density.

Time-lapse continuity inside one fixed composition

  • A locked frame makes temporal consistency easier to judge because camera motion is not hiding drift.
  • Name what must stay fixed and what is allowed to change across weather, season, or time.

Reference motion clip

Macro product lighting and material response

Applicable scenario:

For a product hero shot, describe the object as fixed and let the environment carry heat shimmer, sparks, reflections, glowing material, or particles.

Macro product lighting, reflections, and material effects

  • Macro scenes reveal reflections, heat shimmer, particles, and edge highlights better than generic hero shots.
  • Keep the object structurally stable while the environment carries most of the movement.

Reference motion clip

Fast chase continuity and camera banking

Applicable scenario:

For fast chase or FPV clips, establish the path and camera behavior before styling. Banking, acceleration, blur, and lighting pulses should support one direction.

High-speed camera banking and environment streaks

  • The path, camera angle, lighting pulses, and foreground blur all need to agree.
  • Anchor the setting first, then add acceleration and banking so continuity errors are easier to spot.

Reference motion clip

Object interaction and rhythmic motion

Applicable scenario:

For music, craft, or tactile object videos, name the repeated action and its cadence so hands, tools, surfaces, and contact stay coherent.

Repeated object interaction and rhythmic performance cues

  • Repeated action is the test: hands, tools, surfaces, and rhythm must stay coherent.
  • Describe cadence and contact, not just mood, when prompting craft or performance scenes.

Reference motion clip

Continuous impossible camera move

Applicable scenario:

For one-shot camera journeys, write movement as a sequence of verbs: begins, accelerates, crosses a boundary, breaks through, and continues.

Continuous camera transition across water, light, and scale

  • Look for a continuous camera path across different density, light, and scale conditions.
  • Use clear verbs in the prompt so the move has a beginning, crossing point, and continuation.

Reference motion clip

Product silhouette lock with material morph

Applicable scenario:

For product variations, hold the shape, proportions, and key details steady while the surface changes.

Product silhouette lock during material morphing

  • Material morphing is useful only if the product silhouette stays fixed while the surface changes.
  • Watch laces, edges, and outlines; drifting details need stronger identity-preservation language.

Reference motion clip

Human movement and portrait continuity

Applicable scenario:

For a person-centered video, specify body action first: step, turn, hand raise, torso pivot, expression change, and clothing response.

Human gesture, full-body movement, and fabric inertia

  • Human tests need body mechanics: footwork, torso rotation, arm hits, expression, and fabric response.
  • Start the person in motion immediately and keep the camera readable enough to inspect gestures.

Reference motion clip

Animal behavior and small prop action

Applicable scenario:

For an animal video, make the action concrete and small: the bird reaches, swipes, freezes, hides the phone, and settles back.

Animal behavior, object interaction, and small expressive motion

  • Check whether anatomy and behavior stay readable while comedy or prop interaction is added.
  • Keep the action beat simple so tiny movements remain visible instead of becoming noisy.

Analysis

Seedance 2 capability breakdown for Veyloom

Text / image / video / frame inputs

Multimodal reference planning

Official material emphasizes multimodal references; Veyloom makes text, image, video, and frame-transition inputs available through its creation tools.

Prompt layer

Camera language

Treat camera direction as its own instruction layer: shot size, path, speed, and whether the move reveals, follows, or holds the subject.

Source image plus prompt

Reference-led animation

Image-to-video is the cleanest entry point when visual identity is already decided and the prompt only needs to guide motion.

Source video plus prompt

Source-motion edits

Video-to-video works better when the source clip already has timing or action that should guide the result.

Source image frames

Transition control

Frame transition is useful when the first and last visual states matter more than open-ended exploration.

Workflow-specific

Audio boundary

Generated audio is available for text-to-video and image-to-video here; it is not universal across every Seedance workflow.

Fit

What it is good at

Prompt-led motion

Try short prompt-led motion drafts when the brief depends on subject, action, camera direction, and lighting.

Reference image animation

Use image-to-video when a still image should guide the visual reference for a short clip.

Source-video guidance

Use video-to-video when an existing clip should guide a visual restyle or source-motion edit.

Frame-to-frame planning

Use frame transition when the job needs motion between source image frames rather than a text-only start.

Product

Use it on Veyloom

Seedance workflows on Veyloom

These rows reflect the options currently available in Veyloom's live generator. Check the generator before creating.

Try Seedance 2 for Image to Video

Supported workflow

Image to video

seedance-2-standard-i2v, seedance-2-fast-i2v

Available variant

Seedance 2.0 / Seedance 2.0 fast

Supported input type

Source image plus prompt

Confirmed settings

  • Audio: Supported
  • Ratio: 16:9 / 9:16
  • Duration: 5s / 10s / 15s
  • Resolution: 480p / 720p

Supported workflow

Text to video

seedance-2-standard-t2v, seedance-2-fast-t2v

Available variant

Seedance 2.0 / Seedance 2.0 fast

Supported input type

Text prompt

Confirmed settings

  • Audio: Supported
  • Ratio: 16:9 / 9:16
  • Duration: 5s / 10s / 15s
  • Resolution: 480p / 720p

Supported workflow

Video to video

seedance-2-video-to-video

Available variant

Seedance 2.0

Supported input type

Source video plus prompt

Confirmed settings

  • Audio: Unavailable here
  • Ratio: 16:9 / 9:16
  • Duration: 5s
  • Resolution: 480p / 720p

Supported workflow

Frame transition

seedance-2-frame-transition

Available variant

Seedance 2.0

Supported input type

Source image frames plus prompt

Confirmed settings

  • Audio: Unavailable here
  • Ratio: 16:9 / 9:16
  • Duration: 5s
  • Resolution: 480p / 720p

Choose

Which Seedance workflow should you start with?

Supported workflow

Text to video

Use when

You know the subject, action, camera, and lighting but do not have source media.

Best for

Starting from an idea, tagline, or scene brief.

Watch out for

Needs stronger prompt detail because there is no visual reference.

Supported workflow

Image to video

Use when

You already have a still image and want motion while preserving the visual direction.

Best for

Animating product frames, concept art, character stills, or campaign key visuals.

Watch out for

Too much motion can break the reference; start with a short, concrete move.

Supported workflow

Video to video

Use when

You want an existing clip to guide motion, timing, or composition.

Best for

Restyling a source clip or keeping action timing from a previous take.

Watch out for

The source video drives the result, so weak input footage limits the edit.

Supported workflow

Frame transition

Use when

You can define the starting and ending visual states.

Best for

Planned before/after moments, product transformations, or scene transitions.

Watch out for

Works best when the two frame states are visually compatible.

Prompt diagnosis

Turn vague video prompts into controllable Seedance instructions

Weak prompt

Make this product look cinematic with cool motion.

Why it is weak

It asks for a vibe but gives no subject hierarchy, camera path, or motion boundary.

Veyloom-ready version

A glass perfume bottle on wet black stone, slow push-in camera, thin mist behind the bottle, cool rim light, no label distortion.

Weak prompt

Animate this room and make it beautiful.

Why it is weak

It does not say which parts of the reference image must remain stable.

Veyloom-ready version

Use the uploaded interior image as the layout reference. Add a slow left-to-right dolly, light curtain motion, warm lamp flicker, and stable furniture positions.

Weak prompt

A stylish person moves dramatically in a movie scene.

Why it is weak

It combines identity, action, and style without protecting anatomy or gesture continuity.

Veyloom-ready version

A person turns toward a window, coat fabric moves with the turn, medium shot, soft daylight, natural hands, no face morphing.

Decision notes

Where Seedance 2 on Veyloom fits, and where to be careful

Good fit

  • Reference-led image-to-video when the still frame already carries the art direction.
  • Short motion tests that need clear camera and environment instructions.
  • Workflows where audio is useful but should be checked per selected mode.
  • Frame transition planning when the start and end states are known.

Watch-outs

  • Unbounded prompts that ask for a whole film instead of one controlled clip.
  • Claims of official benchmark parity; official examples are external references only.
  • Video extend on this page; it is not one of the Seedance 2 options shown here.
  • Any assumption that every Seedance workflow supports audio.

Prompts

Prompt Patterns

Original short patterns for audio-capable Seedance text-to-video and image-to-video workflows. They are not official prompts.

SubjectA matte-black product prototype on a mirrored table.

MotionThe object rotates once while fine particles drift outward.

CameraSlow lateral dolly, no abrupt cuts.

LightingControlled studio key light with a narrow cyan rim.

AudioSoft room tone with one clean mechanical accent.

SubjectA quiet city corner after a brief storm.

MotionA cyclist passes through the frame as rain ripples in puddles.

CameraHandheld push-in that settles into a locked close-up.

LightingLate evening practical lights reflected on wet pavement.

AudioLow ambient street texture, no dialogue.

SubjectA paper craft scene built from layered white cardstock.

MotionPaper layers lift and fold into a compact architectural form.

CameraTop-down move into a 45-degree editorial angle.

LightingSoftbox highlights with warm shadow falloff.

AudioSubtle cloth movement and distant studio air.

Notes

Practical notes and limitations

Outputs can vary by prompt, input media, selected settings, and service availability.

Official source examples are external references, not controlled Veyloom benchmarks.

Reference clips are attributed context clips, not generated-on-Veyloom examples.

Current Veyloom options differ by workflow and variant.

Check current options in the live generator before creating.

Reference

FAQ, sources, and related links

What is Seedance 2 on Veyloom?

Seedance 2 on Veyloom connects official ByteDance Seed source links with Seedance workflows currently available on Veyloom.

Which Seedance workflows are currently available on Veyloom?

The Seedance workflows shown here are text to video, image to video, video to video, and frame transition.

Can I use Seedance 2 for image-to-video on Veyloom?

Yes. Veyloom offers Seedance 2.0 image-to-video options that accept a source image plus prompt.

Does the current Veyloom workflow support audio?

Text-to-video and image-to-video Seedance options support generated audio. Video-to-video and frame transition do not support audio here.

Are the examples on this page generated on Veyloom?

The local motion-study clips are Veyloom sample assets. Reference clips are attributed separately, and official examples remain external links only.

Create

Start from the live image-to-video generator.

The CTA opens Veyloom's existing image-to-video workspace with Seedance 2.0 selected.

Try Seedance 2 for Image to Video