Seedance NSFW Prompt Guide for mybabes.ai
05 STEPS
· 10 MIN READ
Intro
Seedance on mybabes.ai is currently one of the most capable unrestricted video models available for adult content. It can generate realistic skin, bodies, fluids, clothing removal, intense physical action, facial expressions, and multi-shot sequences with a level of control that most censored models simply cannot match.
But raw power alone is not enough. The quality of your results depends almost entirely on how you direct the model.
This guide treats Seedance like a professional movie-making tool rather than a simple “type something sexy and hope” generator. It combines two proven systems:
• The Director method — a precise, film-style approach that forces the model to react to what can actually be seen and measured (body position, light direction, muscle movement, camera behavior, physics).
• The official BytePlus Dreamina Seedance 2.0 prompting guidelines — the native multimodal reference system, shot sequencing, asset handling, and consistency techniques designed specifically for this model family.
When these two approaches are used together, you move from random lucky generations to deliberate, controllable adult scenes. You can lock a character’s face and body, design exact blocking, control camera language, extend clips cleanly, and stitch them into longer shorts or full scenes.
The rest of this guide will show you exactly how.
Seedance
Seedance 2.0 (the version running unrestricted on mybabes.ai) is a multimodal video model built by ByteDance. It generates both video and audio in one pass and accepts multiple types of input at the same time:
• Character reference images (face + body)
• Scene / environment images
• Optional motion or camera-reference videos
• Optional audio clips for voice or atmosphere
Because it is unrestricted on mybabes, it will follow explicit sexual direction without refusal — oral, mutual masturbation, grinding, clothing removal, fluids, rough handling, multiple partners, specific fetishes, and most forms of penetration. The limiting factor is almost never the model’s willingness. It is the precision of your instructions.
Important limitation
Seedance is currently weaker at fully explicit, close-up, high-detail penetration (especially deep insertion with clear anatomical accuracy and continuous thrusting). For those specific shots, results can look soft, distorted, or anatomically incorrect.
When you need strong, clear, realistic penetration, switch to the Wan model or the dedicated mybabes video model and follow their specific guides instead. Use Seedance for everything else: teasing, oral, mutual play, body worship, facial reactions, multi-shot storytelling, and longer continuous scenes.
Seedance reacts strongly to visible, measurable details. It is weak with vague mood words (“sexy”, “hot”, “passionate”, “intense”). It is strong with concrete physical direction: exact body positions, how skin deforms, where hands go, how light hits wet surfaces, the speed and force of movement, and clear camera behavior.
It also has native strengths that matter for adult work:
• Good identity lock when face + body references are used correctly
• Strong performance on continuous single takes (ideal for intimate or POV scenes)
• Decent multi-shot sequencing when you structure the prompt properly
• Built-in audio generation (moans, wet sounds, dialogue, ambient noise)
• Extension capability so you can continue a clip instead of starting from zero every time
Used casually, Seedance produces average porn. Used with the methods in this guide, it becomes a high-control adult filmmaking tool — just know when to switch models for the most explicit penetration shots.
Movie Making Tool
Think of Seedance less as a simple “AI video button” and more as a virtual film crew that only follows clear direction.
A real director does not say “make it hot.” They decide:
• Exact body positions in the first frame
• How the performers move and interact
• Where the camera sits and how it moves
• What the light is doing on skin and wet surfaces
• The emotional and physical progression of the scene
That same mindset produces the best results with Seedance.
When you treat it as a movie-making tool you gain:
• Consistent character identity across multiple shots
• Deliberate camera language (slow push-ins, low angles, tracking shots, locked close-ups)
• Much stronger physical realism (skin deformation, fluids, fabric, weight, contact)
• The ability to plan multi-shot sequences instead of one random clip
• Clean extensions so you can build longer scenes and short-form content
Seedance is excellent for teasing, oral, mutual masturbation, body focus, facial performance, grinding, and most forms of intimate interaction. For the most explicit close-up penetration with high anatomical accuracy, switch to Wan or the dedicated mybabes video model as noted earlier.
How this video was created: (first 15 seconds)
Video prompt:
SCENE CONTEXT
15-second 9:16 vertical English pre-roll. Very fast-paced, fun and explicit. Starts with a comedic step-bro slap, then rapid transformations: dragon ride with heavy breast bounce, cowboy outfit with real moan, and movie studio directing moment. All spoken dialogue MUST be in clear English with visible mouth movement. Young early-to-mid 20s woman with full natural heavy breasts. Warm cinematic lighting, no neon. Lots of effects and energy. Vertical framing.
ACTIVE REFERENCES
None. Young woman in her early-to-mid 20s with full natural heavy breasts. Same face across all live-action transformations.
LOCATION MAP
Modern bedroom → fantasy sky with dragon → western desert → professional movie studio. Warm cinematic lighting. No neon.
FIRST FRAME / BLOCKING
Vertical 9:16. Young fully nude (or barely covered) platinum-blonde woman looking at camera with a teasing, playful expression, already mid-energy. Framed tightly for vertical.
FORMAT MODE
Timed multishot with rapid hard cuts for fast pace. Camera does not add extra cuts.
OPTICS
29°–35° portrait FOV, strong motion blur on fast movement, shallow depth on faces, no drift mid-segment.
CAMERA
Very energetic handheld, quick push-ins and short dynamic moves that always emphasize the breasts and face. Vertical framing stays tight on the body.
ACTION
0.0s–3.2s — Step-bro slap
Young woman looks at camera and says “Heeey step-bro…”. A second young woman enters fast and firmly slaps her. The second woman immediately says “NO! Not this again. Let’s create some proper porn.” Clear mouth movement and energy.
3.2s HARD CUT
3.2s–6.8s — Dragon
Same young woman (nude or very revealing) is riding a dragon through a fantasy sky. Her full heavy breasts bounce hard and freely with the flight. Soft magical particles and wind effects. She looks at camera and speaks with visible mouth movement.
6.8s HARD CUT
6.8s–11.0s — Cowboy
Same young woman now in a sexy open cowboy outfit (hat, open shirt) in a western desert. She speaks the next lines, then lets out a clear, real, filthy moan while her breasts move. Soft dust and light effects.
11.0s HARD CUT
11.0s–15.0s — Movie studio
Same young woman standing in a professional movie studio (lights, camera, director feeling). She looks straight at camera with confident energy and delivers the final line while her breasts have natural movement. Soft light flares and particles.
PERFORMANCE
Young, playful, comedic on the slap, then openly sexual and confident. Strong direct eye contact. Clear visible talking in English on every segment. High energy, slightly breathless trailer feel.
PHYSICS
Very prominent heavy natural breast mass. Especially strong, free bounce on the dragon ride. Natural movement in cowboy and studio scenes.
LIGHTING
Warm cinematic throughout with soft particles and light effects. No neon.
AUDIO
Clear energetic young female English voices (must sound natural and spoken in English) + driving sexy background music + rich sound design:
Dialog (all in English, fast and punchy):
“Heeey step-bro…”
“NO! Not this again. Let’s create some proper porn.”
“You can start creating your own porn.”
“Add moaning, dirty talk, sound effects…”
(real moan)
“Boom. You’re directing your own little porn movie.”
Sound effects:
Sharp slap sound
Strong whoosh + light magical chime on every hard cut
Soft dragon wing and wind sounds
Soft skin and heavy breast bounce sounds
Soft western ambient + dust
Soft studio light flares
Low sensual ambient pulse + rising energy
STYLE
Photoreal adult, young glamorous woman, very strong breast physics, cinematic 9:16 vertical, fast, fun, high-energy with clear effects and transformations.
OUTPUT SETTINGS
Real-time, 9:16 vertical, high detail.
POSITIVE LOCKS
All spoken dialogue is clearly in English with visible mouth movement. Fast pace with energetic cuts. Strong breast bounce on the dragon. Clear real moan in cowboy scene. Same young woman across all shots. Early-to-mid 20s. No neon. Vertical 9:16 framing.
How to Craft the Prompt Correctly
The difference between average and strong Seedance results is almost always the prompt structure. Follow this order. Use only the blocks you need. Keep language concrete and visual.
1. SCENE CONTEXT
Short and clear. Tell the model the situation and location in one or two sentences. No mood words.
2. ACTIVE REFERENCES
This is where you lock the character. Always declare what each image is for and put the face first.
3. LOCATION MAP
Quickly map the space so the model doesn’t invent random backgrounds.
4. FIRST FRAME / BLOCKING
One of the most important blocks. Describe exactly how the body is positioned when the video starts.
5. FORMAT MODE
Choose continuous or multi-shot.
Continuous (best for most intimate scenes):
Good examples:
• “MCU, 29° FOV. Slow steady push-in.”
• “CU, 18° FOV. Locked off, eye-level.”
• “MS, 47° FOV. Smooth low-angle tracking from her feet toward her face.”
• “ECU, 12° FOV. Static on her hand and the wetness between her legs.”
7. ACTION
Describe the physical movement in clear, measurable terms.
8. PERFORMANCE
This is where most of the erotic charge comes from. Be specific about face, eyes, breathing, and sounds.
9. PHYSICS
Tell the model how bodies and fluids actually behave.
10. LIGHTING
Light reveals form and wetness. Side light or soft window light usually works best.
11. AUDIO (optional but powerful)
• Write what the camera sees, not how the scene “feels”
• First frame must be clear and specific
• Prefer slower, continuous movements
• One main camera idea per shot
• Always lock identity and add the anti-text / anti-watermark lines
How this video was created: (first 15 seconds)
SCENE CONTEXT
15-second 9:16 vertical English pre-roll. Extremely fast-paced, fun and explicit. Starts with dramatic STOP sign in the storm, then rapid cuts of different young women. Special care on the “Create me / Or me / Or us” sequence so each line matches the correct woman. All spoken dialogue MUST be in clear English with visible mouth movement. Young early-to-mid 20s women with full natural heavy breasts. Warm cinematic lighting, no neon. Vertical framing.
ACTIVE REFERENCES
None. Young women in their early-to-mid 20s with full natural heavy breasts.
LOCATION MAP
Stormy outdoor night with heavy sideways rain and strong wind → warm modern interiors that change quickly. No neon.
FIRST FRAME / BLOCKING
Vertical 9:16. Fully nude young woman with massive heavy breasts stands in a violent sideways storm, both hands gripping a red STOP sign pulled by the wind. She looks straight at camera with intensity. Framed tightly for vertical.
FORMAT MODE
Timed multishot with many rapid hard cuts. Camera does not add extra cuts.
OPTICS
29°–35° portrait FOV, strong motion blur on wind and body movement, shallow depth on faces, no drift mid-segment.
CAMERA
Very energetic handheld, quick push-ins that always emphasize the breasts and face. Vertical framing stays tight on the body.
ACTION
0.0s–3.2s — STOP sign
Fully nude young woman fights strong wind and rain while gripping the red STOP sign. Hair and huge breasts whip and bounce hard. She looks at camera and firmly says the opening lines with clear mouth movement.
3.2s HARD CUT
3.2s–5.8s — First woman (“Create me”)
Young platinum-blonde fully nude woman walks toward camera with strong breast bounce. She looks at camera and clearly says “Create me…”
5.8s HARD CUT
5.8s–8.4s — Second woman (“Or me”)
Young dark-haired fully nude woman in a seductive pose, one hand under her breast. She looks at camera and clearly says “Or me…”
8.4s HARD CUT
8.4s–11.0s — Third woman (“Or her”)
Young light-brown-haired fully nude woman turning toward camera with clear breast movement. She says “Or her…”
11.0s HARD CUT
11.0s–15.0s — Two women together (“Or us”)
Blonde + dark-haired women standing close together, looking at camera. They say together: “Or us. Any girl you want.”
PERFORMANCE
Young, intense on the storm, then playful, confident and sexy. Strong direct eye contact. Clear visible talking in English. Each “Create me / Or me / Or her / Or us” is cleanly delivered by the correct woman on her cut. High energy and fast rhythm.
PHYSICS
Very prominent heavy natural breast mass. Extreme bounce and whip from the wind on the first shot. Strong natural bounce on every following cut.
LIGHTING
Cold dramatic storm light with rain streaks → warm soft cinematic light in interiors. No neon.
AUDIO
Clear energetic young female English voices (must sound natural and spoken in English) + driving sexy background music + rich sound design:
Dialog (precisely timed to the cuts):
“STOP watching porn!
Create your own porn instead.
Create me…”
“Or me…”
“Or her…”
“Or us.
Any girl you want.”
Sound effects:
Strong wind and heavy rain on opening
Sharp whoosh on every hard cut
Soft skin and heavy breast bounce sounds
Low sensual ambient pulse + rising energy
STYLE
Photoreal adult, young glamorous women, very strong breast physics, cinematic 9:16 vertical, extremely fast, fun and highly dynamic with many cuts.
OUTPUT SETTINGS
Real-time, 9:16 vertical, high detail.
POSITIVE LOCKS
Opens with fully nude woman holding STOP sign in strong wind/rain, huge breasts bouncing. The lines “Create me”, “Or me”, “Or her”, “Or us” are each spoken by the correct woman on her specific cut with clear mouth movement. All dialogue in English. All women early-to-mid 20s with prominent heavy breasts. No neon. Vertical 9:16 framing.
Good and Bad Results
Learning to recognise the difference between a weak generation and a strong one will save you a lot of tokens and time.
Bad Results (Common Failures)
• Face drifts or changes mid-video
• Body proportions shift or become distorted
• First frame is empty, awkward, or random
• Camera moves in multiple conflicting ways
• Action feels floaty or physically wrong
• Wetness and contact look fake or missing
• Unwanted text, subtitles, logos or watermarks appear
• Character loses eye contact or looks in the wrong direction
• Movement is too fast or broken
• Lighting is flat and hides form
Good Results (What You Want)
• Face and body stay consistent with the references the entire time
• Clear, intentional first frame
• Real weight, skin deformation and fluid behavior
• One controlled camera idea
• Visible wetness and contact
• Strong facial performance
• No text, logos or watermarks
• Smooth continuous movement
• Light that reveals body and wetness
Full Example of a Strong Prompt (Good Result)
SCENE CONTEXT
Bedroom at night. Woman alone on the bed, on her back, masturbating while looking at the camera.
ACTIVE REFERENCES
Reference the woman in @Image1 (face close-up) and @Image2 (full body) as Subject 1. 100% matches the references.
LOCATION MAP
Foreground: white sheets and her lower body. Midground: her torso and face. Background: soft window light and dark wall. Camera is positioned at the foot of the bed, slightly low.
FIRST FRAME / BLOCKING
She is on her back, knees bent and open, one hand between her legs with fingers resting on her clit, the other hand on her breast. She is looking directly into the lens.
FORMAT MODE
one continuous shot, the camera does not cut on its own.
OPTICS
MCU, 29° FOV.
CAMERA
Slow steady push-in from medium close-up toward her face and upper body.
ACTION
Her fingers start with slow circles over her clit, then two fingers slide inside. Her hips roll gently against her hand. The other hand squeezes her breast and pinches the nipple. Wetness increases and becomes clearly visible on her fingers and between her legs.
PERFORMANCE
Eyes locked on the camera the entire time. Mouth slightly open. Soft moans with every breath. Chest rises and falls harder. A flush spreads across her neck and chest. Occasional sharper inhale when her fingers go deeper.
PHYSICS
Natural skin deformation where her fingers press. Realistic wetness and shine on her fingers and between her legs. Soft bounce of her breasts as she moves. Hair slightly messy on the pillow.
LIGHTING
Warm side light from the window (around 3200K), soft fill from the opposite side. Clear sheen on wet skin and fingers. Soft shadows that still show body shape.
CONSTRAINTS
Keep it subtitle-free. Avoid generating any text or subtitles.
Do not generate a logo or watermark.
Face and body remain 100% consistent with the references.
Skin tone, hair and proportions stay locked.
She stays on her back the entire time.
Help Yourself Using Grok or LUNA with the Director Skill
Best Way to Do It
1. Copy the entire Director skill text (the big block below).
2. Paste it into a new chat with Grok or LUNA.
3. Also upload your reference image(s) (face + full body recommended).
4. Clearly tell it what you want to achieve.
Recommended message after pasting the skill:
I just gave you the Seedance Director skill.
Here are my reference images.
Now use the full system and write a complete standalone Seedance prompt for this scene:
[describe exactly what you want in plain language — position, action, camera, mood, specific details]
Why this works better
The skill tells Grok/LUNA how to structure the prompt.
Your reference images give it the exact character identity.
Your description tells it what should happen in the scene.
Full Director Skill (copy everything below)
Description
Seedance 2.0 prompt-writing system. Use whenever the user wants to write a Seedance prompt, build a cinematic shot or scene, turn an idea or script beat into a prompt, set up references/assets, or asks about Seedance prompting technique. Trigger on phrases like “prompt for Seedance”, “write a Seedance prompt”, “make a scene where…”, “build a shot”, “how to write a Seedance prompt”. Output is a single standalone prompt. Always consult this skill before writing any Seedance prompt.
Seedance 2.0 Prompt Writer
Turn scripts, scenes, and ideas into Seedance 2.0 prompts. Block, light, and pace each shot like a film.
Output: always a single standalone prompt in a code block.
CORE PRINCIPLE: WRITE THE VISIBLE
The model reacts to what can be seen and measured, not to mood words.
Translate every abstraction into something observable.
* ❌ “tense scene” → ✅ “man freezes, slowly clenches his fist, light only from the side, half his face in shadow”
* ❌ “cool cinematic shot of a car, epic, fast” → ✅ “low tracking shot alongside the car as it powers through a wet curve, headlights glowing, spray off the tyres, hard buffeting camera shake”
Write in plain, clear, instruction-style language. Fewer precise words beat many vague ones.
Before generating, mentally “watch” the prompt as a viewer: is everything unambiguous, is the first frame non-empty, is it clear where the subject is and where it looks, where is the light coming from.
WORKFLOW
1. Read as a director. Find the dramatic shape: where the scene turns, lands, breathes.
2. Define continuity anchors. Who is in frame, how they look, what they carry across cuts.
3. Write the prompt using the block structure below.
4. Tag assets. Tagging rules below.
If the idea is ambiguous, close the gaps in conversation first — blocking, angle, light, timing, character behavior, what is in the first frame, how it ends. The model fills any gap on its own, usually not the way intended.
ASSET TAGGING
Reference image (@tag) sets appearance and identity. Text sets what happens and locks critical details. Both are required.
Tag naming:
* If the user specifies tags as @tag_name, use those.
* If no tags are specified, name them by load order of the supplied assets: @image1 @image2 … for images, @video1 @video2 … for videos, @audio1 @audio2 … for audio.
* Otherwise follow the user’s explicit instruction.
Reference + text rule: keep the character description minimal — long appearance text conflicts with the image and degrades it.
@TAG: age + role/build + current state + unique visible features + action-critical details + voice (only if it has a line). 100% matches the reference.
State critical details (small text, logos, color) in words even if they appear on the reference — the model can drop them.
Never place an @tag in a shot where that object is not present; the model will try to force it into frame.
CONTEXT ISOLATION
Each generation is a blank slate — no memory of previous shots. A prompt is a sealed single-shot document.
Do not carry in: scene numbers, script headings, summaries of prior scenes, unused tags/characters, “as above / continues” phrasing, or people and props from a previous line that are not needed in this shot.
STYLE — DISTRIBUTED, NOT A PREFIX
There is no separate style-prefix block at the top of the prompt. Style is not one thing in one place — it is a set of aspects, and each aspect lives in the block that already governs it, where it is most effective and unambiguous. State the target look, never what to avoid; adapt every line to this specific scene.
Cover these aspects, each in its home block:
| Aspect | Lives in | What to write |
|--------|----------|---------------|
| Lighting | LIGHTING | source, direction, exposure, key/fill, haze for THIS scene |
| Color / grade | COLOR GRADE (own block if grade is strong) or folded into LOCATION + LIGHTING | palette as material + light beam + role, never a flat list |
| Lens / optical character | OPTICS | FOV°, rectilinear/anamorphic, prime-lens character, motion-blur |
| Camera body / tonal character | CAMERA | tonal latitude, highlight roll-off, color science — as a look, not a model name |
| Skin / micro-realism | PERFORMANCE (or inside ACTION) | pore-level realism, capillary flush, living eyes, catch-lights, visible breath |
| Acting | PERFORMANCE (or inside ACTION) | micro-pauses, precise eye-line, restraint, muscle-level emotion |
| Physics realism | PHYSICS | gravity, inertia, mass, contact shadows, fluids, particles |
| Composition / blocking | FIRST FRAME / BLOCKING | framing rule for THIS scene; everyone moving from frame one |
| Continuity | POSITIVE LOCKS | characters, props, environment identical across cuts |
| Wardrobe | WARDROBE (own block when costume matters) | material + condition, scene-logical |
| Format / resolution / grain / fps / bitrate | STYLE + OUTPUT SETTINGS at the end | 8K, photoreal, anamorphic, fine grain, real-time vs slow-mo per segment |
PROMPT STRUCTURE
Write blocks in this order. Use only the blocks the shot needs.
SCENE CONTEXT
[1–2 sentences: what happens, where, when. Geo-positions of characters.]
ACTIVE REFERENCES
[@tag + minimal anchor of critical details + “100% matches the reference”]
LOCATION MAP
[Foreground / midground / background, where the camera is, where light comes from, movement paths.]
FIRST FRAME / BLOCKING
[Who is where in the first frame: positions, orientation, gaze. Composition rule for this scene.]
FORMAT MODE
[one continuous shot, the camera does not cut on its own.
OR CUT 1 … CUT 2 … CUT 3
OR timed HARD CUTs]
OPTICS
[Shot size + FOV per segment + lens/optical character. For multishot add “no drift mid-segment”.]
CAMERA
[Operator behavior: height, distance, movement, focus.]
ACTION
[Events at the precision the shot needs. Camera motion and subject motion stated separately.]
PERFORMANCE
[Acting + skin/micro-realism: muscle-level emotion, eye-line, catch-lights, breath, pore-level detail.]
PHYSICS
[Mass, inertia, contact, fluids, particles.]
LIGHTING
[Source, direction, exposure — priority block.]
COLOR GRADE
[Only if stylized]
WARDROBE
[Only when costume matters]
AUDIO
[Only the needed sound / line.]
STYLE
[Overall look in words, photoreal, format, grain]
OUTPUT SETTINGS
[Resolution, anamorphic, real-time vs slow-mo if needed]
POSITIVE LOCKS
[Short fixers against likely failures, in positive form]
FINAL RULES
- English prompts only.
- Positive phrasing only.
- No director names, no equipment names.
How to Stitch Videos and Create Shorts (Extend Feature)
Seedance on mybabes.ai has a built-in Extend function. This lets you continue an existing clip instead of starting from zero every time. Used correctly, it is the fastest way to build longer scenes and short-form content.
How to Stitch Videos and Create Shorts (Extend Feature)
When to Use Extend vs New Generation
• Use Extend when you want to continue the same action, same camera angle, or same emotional beat (especially good for teasing, oral, mutual play, or slow build-ups).
• Start a new generation when you need a clear change in angle, position, or intensity (for example going from oral to penetration, or switching from wide to extreme close-up).
Best Practices for Clean Extensions
1. Generate a strong first clip using the full Director structure.
2. When extending, keep the same character references.
3. Tell the model what should continue and what should change.
4. Prefer continuing the same camera language rather than suddenly changing it mid-extension.
5. Keep movements relatively slow and continuous — abrupt high-energy actions are harder to extend cleanly.
Example extension request style:
Use last frame of reference video and extend the video using following:
Stitching Tips for Smooth Results
• Try to match lighting and body position as closely as possible between clips.
• If the join still feels rough, add a very short crossfade or a quick cut on a strong movement (hand movement, hip roll, head turn).
• Keep the same character references across all clips in one short so identity stays locked.
Continuous Long Take vs Segmented Shorts
• Long continuous take (heavy use of Extend) → best for intimate, hypnotic, POV, or slow-burn scenes.
• Segmented shorts (several separate generations stitched) → best for higher energy, multiple angles, or clear progression (tease → action → climax).
Most creators get the best results by mixing both: start with a strong continuous take, extend it once or twice, then cut to a new angle for the peak moment.
Final Advice
Always generate the first clip with a full structured prompt.
Weak first clips produce weak extensions.
Strong first clips + careful extending + simple stitching = consistent, high-quality short-form adult content.
A good prompt is a series of clear choices.
Keep this guide beside your next session. Choose one layer, write it plainly, run the result, and improve only what needs direction.
Start a new prompt