Seedance prompts: how to write multi-shot video prompts

The multi-shot structure

A single-shot prompt describes one moment. A multi-shot prompt describes a small sequence, so it needs an extra layer: continuity. This is the structure that works for most scenes:

1. Setup line: character (age, clothes, one distinctive detail), place, time of day, overall look.
2. Shot lines: "Shot 1:", "Shot 2:" and so on, each with shot size, camera move and one action.
3. Continuity line: "Same character, same outfit, same light in every shot."

Some Seedance versions handle several shots in one generation and some work better one shot at a time. If yours does not cut cleanly, generate each shot separately and paste the same setup line into every prompt. Check the model's current limits for length and shot count.

6 Seedance prompt examples with Indian scenes

1. Morning rush, Pune hostel

A 20-year-old engineering student with curly hair, round glasses and a grey hoodie, in a small hostel room in Pune, morning, soft natural light, realistic.
Shot 1: Wide shot, she wakes up to a phone alarm and sits up.
Shot 2: Close-up, she squints at the phone screen and her eyes go wide.
Shot 3: Medium shot, tracking, she runs down the hostel corridor pulling on her backpack.
Same character, same hoodie and glasses in every shot.

Why it works: three shots, three sizes, one clear story (late for class). The glasses and grey hoodie are simple anchors the model can repeat.

2. Kirana store conversation

A friendly kirana shop owner in his fifties with a grey moustache and a checked shirt, and a young woman customer in a blue kurta, inside a small neighbourhood store in Indore, warm tube light, shelves full of packets.
Shot 1: Over the shoulder of the customer, the shop owner smiles and says, "Aaj kya chahiye, beta?"
Shot 2: Over the shoulder of the shop owner, she replies, "Bas ek packet chai aur doodh."
Shot 3: Close-up, he hands her a paper bag across the counter.
Same characters, same clothes, same light.

Why it works: the over-the-shoulder pair is the classic dialogue pattern, and each shot has one short line. Naming both characters in the setup keeps them from swapping looks.

3. Handmade candle product reveal

A small handmade soy candle in a matte terracotta pot, on a wooden table by a window in a Jaipur studio, late afternoon golden light, calm premium look, no text, no logo.
Shot 1: Extreme close-up, a match strikes and flares.
Shot 2: Close-up, the flame touches the wick and catches.
Shot 3: Medium shot, slow pull out to show the lit candle, a cup of tea and an open book on the table.
Same candle and pot in every shot.

Why it works: it builds from detail to context, which is how product ads usually cut. The warm light is set once and carried through.

4. Evening on the Varanasi ghats

The ghats of Varanasi at dusk, warm lamplight, a calm and respectful documentary mood.
Shot 1: Wide shot from a boat on the river, the steps glowing with small oil lamps.
Shot 2: Medium shot, an elderly boatman in a white vest rows slowly, oars dipping into the water.
Shot 3: Close-up, a single leaf boat with a small flame floats past on the dark water.
Same evening light and colour in every shot.

Why it works: it shows a sacred place quietly, without staging rituals or inventing anything. If you make devotional content, check that the imagery suits the tradition and your audience.

5. Rainy bus stop in Shillong

A young man in his twenties with a mustard scarf and a black umbrella, and a young woman in a green raincoat, at a small bus stop shelter on a misty hill road in Shillong, light rain, soft grey light, gentle romantic mood.
Shot 1: Wide shot, he waits alone under the shelter as rain falls.
Shot 2: Medium shot, she runs in from the rain, laughing, shaking water off her raincoat.
Shot 3: Close-up, he tilts his umbrella toward her without a word.
Same characters, same clothes, same misty light.

Why it works: a tiny story with a beginning, middle and end. The mustard scarf and green raincoat are colour anchors that stop the characters drifting between shots.

6. Dum biryani reveal, Hyderabad

A home cook in her forties with a maroon dupatta, in a busy Hyderabad kitchen, warm tungsten light, steam everywhere, rich food film look.
Shot 1: Close-up, her hands break the dough seal around a heavy handi lid.
Shot 2: Extreme close-up, steam bursts out as she lifts the lid.
Shot 3: Top-down shot, a ladle pulls up long grains of saffron rice and pieces of meat.
Same hands, same dupatta colour, same light.

Why it works: food sequences live on texture and steam. Each shot moves closer to the reward, and the top-down ending is a natural final frame for a Reel.

Write multi-shot prompts with Real Prompt Generator

The free Seedance prompt generator turns one simple idea into a full prompt with shot size, camera move, light and mood chosen from chips, so you can build each shot fast and keep the look the same. When your sequence comes from a script, Kritiverse's storyboard generator turns the script, Hindi included, into a shot list first.

Tips for cleaner cuts

• Three shots is a good start. More shots in one prompt means more chances for drift.
• Pick anchors the model can see. Colour of clothing, glasses, a scarf, hair style. Not "kind eyes".
• Keep one light source story. Do not move from golden hour to night inside one sequence.
• Vary shot sizes. Wide, medium, close. Two mediums in a row feel like a jump cut.
• Label the final video. Add "Made with AI" and use YouTube Studio's altered or synthetic content setting where it applies.