> ## Documentation Index
> Fetch the complete documentation index at: https://docs.ruapi.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Seedance 2.5 and 2.0 prompts: formula, storyboards, sound and ready-made examples

> How to write prompts for Seedance 2.5 and 2.0: the formula from the official BytePlus guide, storyboards and timestamps, camera movement, sound and dialogue, keeping subtitles out, 10 ready-made prompts and a 480p draft before an expensive 1080p render. Plus the results of our test in Russian.

**In short:** a good Seedance prompt isn't a pretty description — it's instructions for a camera operator: **who** is in the shot and **what they do**, **where** and in what light, **how the camera moves**, the style, and what must not appear. Break complex scenes into shots. Put dialogue in curly braces. You can write the prompt in Russian — we tested it. For dialogue in Russian, choose `doubao-seedance-2.5`. Try a prompt on a cheap 480p clip first, and only then render in 1080p.

The tips below are based on BytePlus's official prompt guides for [Seedance 2.0](https://docs.byteplus.com/en/docs/modelark/seedance-2-0-prompt-guide) and [Seedance 2.5](https://docs.byteplus.com/en/docs/modelark/seedance-2-5-prompt-guide), and on our own test of prompts and dialogue in Russian.

## The prompt formula

The "advanced formula" from the official Seedance 2.0 guide:

> **subject + action + scene + lighting and color + camera movement + style + image quality + constraints**

| Part | What to write | Example |
| - | - | - |
| Subject | Who or what is in the shot: 2–3 consistent, distinctive features | a ginger cat with a blue collar |
| Action | A specific movement, with its speed and intensity | slowly stretches and yawns |
| Scene | Where it all happens | on a windowsill, snow falling outside |
| Lighting and color | Time of day, light source, color palette | soft morning light, warm tones |
| Camera | One movement per shot | medium shot, static camera |
| Style | Overall look | cinematic, film grain |
| Image quality | Sharpness and detail | rich detail, natural colors |
| Constraints | What must not appear | no subtitles and no text on screen |

First pin down who is in the shot and what they're doing. Then where it happens and the mood. Then how to shoot it. Finally, add the style, quality and constraints.

## Which language to write in

Russian isn't on the official list of Seedance 2.5 languages. The list has 11: Chinese, English, Spanish, Indonesian, Malay, Thai, Arabic, Portuguese, Vietnamese, Japanese and Korean ([BytePlus](https://docs.byteplus.com/en/docs/ModelArk/seedance-2-5), November 2026). We tested how the models handle Russian (November 2026, one 4-second 480p clip per case):

| What we tested | `doubao-seedance-2.0-mini` | `doubao-seedance-2.5` |
| - | - | - |
| Prompt in Russian | Followed the description: the cat's color, the collar, snow outside, the action | Followed the description, including the camera movement |
| Dialogue in Russian | Recognizable as Russian speech, but some words are garbled | The line was spoken accurately — whether the prompt was in Russian or English |

Takeaways:

* **You can write the prompt in Russian.** On `doubao-seedance-2.0-mini` we compared the same prompt in Russian and in English and saw no difference in the picture. We haven't tested long, complex scenes in Russian — for those, English is the safer choice.
* **For dialogue in Russian, choose `doubao-seedance-2.5`.** In our test, `doubao-seedance-2.0-mini` garbled some words of the Russian line.
* **State the language of the line:** `says in Russian {…}` (or `говорит по-русски {…}` if the whole prompt is in Russian). The official 2.0 guide recommends marking all dialogue that isn't in Chinese or English this way. Keep all dialogue in one clip in the same language.

## Storyboards and timestamps

If the clip has more than one event, describe them in order — this helps the model keep them in sequence:

```text theme={null}
Shot 1: a young woman walks into a small cafe, the camera smoothly tracks her from the side.
Shot 2: she sits down by the window and opens a book, medium shot, static camera.
Shot 3: she looks up toward the window and smiles, the camera slowly pushes in.
```

* **Seedance 2.0** only understands shot numbers and ignores timestamps. Don't set a duration for each shot.
* **Seedance 2.5** understands both shot numbers and whole-second timestamps, for example `0-3 seconds: … 3-7 seconds: … 7-15 seconds: …` or `[1s-4s] … [4s-8s] …`. The intervals must be continuous, with no gaps. Timestamps don't work for very frequent actions such as "shakes their head three times per second".
* Within each shot, describe in order: camera movement → action and emotion → where the subject is → sound.

## Actions and emotions

* **Be specific, including speed and degree:** not "moves", but "slowly raises a hand", "smoothly turns their head", "slightly lowers their head".
* **Prefer smooth, small movements.** The official guide advises avoiding sudden, large actions such as sprinting, big jumps and rolls.
* **Link actions together:** "turns around and, without stopping, raises a hand".
* **Show emotions through visible details instead of naming them:**

| Instead of | Write |
| - | - |
| she's sad | lowers her head, shoulders trembling slightly, eyes reddening |
| he's happy | he can't help smiling, a spring in his step |
| she's nervous | keeps glancing at her watch, drumming her fingers on the table |
| he's angry | fists clenched, jaw tight, chest heaving |

## Camera

Seedance understands standard terms well: wide shot, medium shot, close-up, push-in, pull-out, pan, tracking shot, orbit, static camera, overhead shot, low-angle shot.

**One camera movement per shot.** If you ask for a push-in, a pull-out and a pan at the same time, the picture becomes less stable.

## Sound, dialogue and subtitles

Seedance 2.0 and 2.5 generate video with sound out of the box. To help the model tell types of information apart, the official 2.0 guide suggests these brackets:

| What | Brackets | Example |
| - | - | - |
| Music | `( )` | `(fast-paced rock music in the background)` |
| Sound effect | `< >` | `<a dog barking in the distance>` |
| Dialogue | `{ }` | `says in Russian {Ты всё-таки пришёл!}` |
| Subtitles | `[ ]` | `[Chapter One: Departure]` |

* If you don't need sound, pass `"generate_audio": false` (inside `metadata` for `/v1/videos`) — see the [video generation guide](https://docs.ruapi.ai/en/video-generation.md).
* The model sometimes misspells on-screen text, swapping or dropping letters. The official 2.5 guide suggests two fixes: ask for the word's letters to appear one by one, in order, or pass the exact text as a reference image.

## Constraints: subtitles, logos, watermarks

Sometimes the model adds subtitles, a logo or a watermark on its own. Add constraints at the end of the prompt:

```text theme={null}
No subtitles and no text on screen. No logos. No watermarks.
```

* You can't rule out subtitles completely, but this makes them less frequent — especially in landscape (`16:9`), where they appear less often than in portrait.
* For **Seedance 2.5**, the official guide recommends describing what should be in the shot rather than listing what shouldn't, where possible. Negative constraints work reliably for subtitles and sound ("no subtitles", "no background music").
* If you want a specific style (anime, 3D cartoon) but the reference image is realistic, state the style explicitly — otherwise the clip may drift toward realism.

## First frame and references

* **Image-to-video (first frame).** Don't describe what's in the picture — the model can see it. Describe **what should happen**: movement, camera, sound.
* **Multiple references.** Refer to each subject the same way throughout the prompt and say where it comes from: "the woman from image 1", "the scene style from image 2". Don't max out the reference limit: the official guide recommends sticking to 4–5 references, which makes it easier for the model to tell what matters.
* **Photos and videos of ordinary people** are fine; celebrities and film or game characters are not. Seedance 2.0 mini and fast have a few caveats — see [Seedance: real person](https://docs.ruapi.ai/en/seedance-real-person-error.md).

## 10 ready-made prompts

Swap in your own subject and adjust the details. All examples are for text-to-video, without references. We tested prompts 5 and 6 exactly as written; the rest are built on the same formula. Prompts 6 and 10 are meant for `doubao-seedance-2.5`.

> [accordion] 1. Product on a table (ad)
>
> ```text theme={null}
> A frosted-glass perfume bottle stands on a wet black stone. The camera slowly orbits the bottle, soft highlights glide across the glass, a light haze rises around it. Studio lighting, dark background, premium ad, rich detail. (calm electronic music) No subtitles and no text on screen.
> ```

> [accordion] 2. Food close-up
>
> ```text theme={null}
> Close-up: a pat of butter is placed in a hot pan, slowly melting and bubbling. Static camera, slightly overhead. Warm kitchen light, rich colors, shallow depth of field. <butter sizzling in the pan> No subtitles and no text on screen.
> ```

> [accordion] 3. Nature
>
> ```text theme={null}
> Dawn over a conifer forest, sunbeams breaking through the morning fog. The camera slowly flies forward over the treetops. Soft golden light, cinematic landscape, natural colors. <birdsong, light wind> No subtitles and no text on screen.
> ```

> [accordion] 4. City at night
>
> ```text theme={null}
> A night street after rain, neon signs reflected in puddles, a taxi slowly drives past. The camera smoothly tracks it from the side at eye level. Blue-purple tones, cyberpunk, film grain. (slow synthwave) <tires swishing on wet asphalt> No subtitles and no text on screen.
> ```

> [accordion] 5. Cat on a windowsill (tested)
>
> ```text theme={null}
> A ginger cat with a blue collar lies on a windowsill, slowly stretches and yawns. Snow is falling outside the window. Medium shot, static camera. Soft morning light, warm tones. Quiet purring. No subtitles, no text on screen.
> ```

> [accordion] 6. A line in Russian (tested, doubao-seedance-2.5)
>
> ```text theme={null}
> A young woman in a yellow sweater sits by the window in a cozy cafe. She looks up, smiles and says in Russian {Ты всё-таки пришёл!}. Medium close-up, slow push-in. Warm soft light. No subtitles, no text on screen.
> ```

> [accordion] 7. Cartoon character
>
> ```text theme={null}
> 3D cartoon: a small robot with a round head and big eyes sees snow for the first time, carefully catches a snowflake in its palm and smiles happily. Medium shot, the camera slowly pulls back. Bright soft colors, animated feature film style. (light cheerful music) No subtitles and no text on screen.
> ```

> [accordion] 8. Anime
>
> ```text theme={null}
> Anime style: a dark-haired young woman with a short haircut stands on a rooftop in the sunset glow, the wind tousling her hair and scarf; she slowly turns toward the camera. Medium shot, static camera. Warm orange and pink tones, hand-drawn 2D animation. <wind> No subtitles and no text on screen.
> ```

> [accordion] 9. Text in the shot
>
> ```text theme={null}
> A pair of white sneakers sits on the asphalt, the camera slowly pushes in. At the end of the clip, the word "RUN" smoothly appears in the center of the frame in large white letters. Minimalist urban style, daylight. (rhythmic music) No other text and no subtitles.
> ```

> [accordion] 10. Timestamps (doubao-seedance-2.5, 8 seconds)
>
> ```text theme={null}
> 0-3 seconds: an aerial view of a city in the morning, the camera slowly descends to an empty square. 3-6 seconds: a dog calmly walks onto the square, the camera tracks it from the side. 6-8 seconds: the dog sits down by a fountain, static camera, wide shot. Natural colors, soft morning light. No subtitles and no text on screen.
> ```

## How to avoid overpaying for test runs

* **Test on short clips: 480p, 4 seconds.** That's about $0.17 with `doubao-seedance-2.0-mini` and about $0.52 with `doubao-seedance-2.5` at the standard price. Current prices are on the [pricing page](https://www.ruapi.ai/pricing?utm_source=docs\&utm_campaign=en-seedance-prompts). Test timestamps and Russian dialogue on `doubao-seedance-2.5` right away: 2.0 doesn't support timestamps, and in our test `doubao-seedance-2.0-mini` got some Russian words wrong.
* **Use draft mode.** It's available only for `doubao-seedance-2.5` and only in the ModelArk format: render a 480p draft first, then turn the one you like straight into 1080p, keeping the same prompt, references and `seed`. See "Draft mode" in the [video generation guide](https://docs.ruapi.ai/en/video-generation.md).
* **Change one thing at a time** — camera, lighting or action — so you know what worked.
* **Failed tasks aren't billed:** if generation fails, the pre-charge is refunded to your balance.

## How to try Seedance via RuAPI

No code: open the "Video generation" page in the console and paste a prompt — see [Seedance online in the browser](https://docs.ruapi.ai/en/seedance-online.md). Through the API:

**1. Sign up**

Sign up at [www.ruapi.ai](https://www.ruapi.ai/?lng=en\&utm_source=docs\&utm_campaign=en-seedance-prompts) — with email, Google or Yandex.

**2. Top up**

Console → **Wallet**: USDT or USDC (TRC20 / BEP20 / Polygon; USDT also on TON), from 5 USDT. No USDT yet? [Buy it with rubles via SBP (Faster Payments System)](https://docs.ruapi.ai/en/buy-usdt.md). Cards from foreign banks (Visa / Mastercard) are accepted too. Video models become available after your first top-up.

**3. Key**

Console → **API Keys** → **Create API key** — you'll get an `sk-...` string.

**4. Your first video**

Send a prompt from the list above — see the example request below. For prompts 6 and 10, set `"model": "doubao-seedance-2.5"`, and for prompt 10 also `"seconds": "8"`. To poll the task status and download the video, see the [video generation guide](https://docs.ruapi.ai/en/video-generation.md).

```bash theme={null}
curl https://www.ruapi.ai/v1/videos \
  -H "Authorization: Bearer sk-your-RuAPI-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "doubao-seedance-2.0-mini",
    "prompt": "A ginger cat with a blue collar lies on a windowsill, slowly stretches and yawns. Snow is falling outside the window. Medium shot, static camera. Soft morning light, warm tones. Quiet purring. No subtitles, no text on screen.",
    "seconds": "4",
    "metadata": { "resolution": "480p", "ratio": "16:9" }
  }'
```

## FAQ

> [accordion] Can I write the prompt in Russian?
>
> Yes. Russian isn't on the official Seedance 2.5 language list, but in our test both `doubao-seedance-2.0-mini` and `doubao-seedance-2.5` followed Russian prompts accurately. We haven't tested long, complex scenes in Russian — for those, English is the safer choice.

> [accordion] Can a character speak Russian?
>
> Yes, with `doubao-seedance-2.5`: in our test the line was spoken accurately. With `doubao-seedance-2.0-mini`, the speech was recognizably Russian, but some words got garbled. Put the line in curly braces and state the language: `says in Russian {…}`.

> [accordion] Why did subtitles or a watermark appear in the video?
>
> The model sometimes adds them on its own. Add "No subtitles and no text on screen. No logos. No watermarks." to the prompt and use landscape format where possible.

> [accordion] Why was the task rejected with SensitiveContentDetected or PolicyViolation?
>
> The prompt or image failed the model's moderation: for example, the prompt names a celebrity or a character from a film or game. The codes are explained in [Seedance: PolicyViolation and SensitiveContentDetected](https://docs.ruapi.ai/en/seedance-sensitive-content-error.md).

> [accordion] Is there a subscription?
>
> No. You pay as you go, only for successfully generated videos. Top-ups start at 5 USDT.

## Next

> [card] [Video generation: Seedance 2.0 / 2.5](https://docs.ruapi.ai/en/video-generation.md)
>
> All modes, the ModelArk format, drafts and prices.

> [card] [Seedance API](https://www.ruapi.ai/seedance-api?lng=en\&utm_source=docs\&utm_campaign=en-seedance-prompts)
>
> Features, prices and payment without a foreign card.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.