Exine AI — Все ИИ в одном месте
24.6K subscribers
54 photos
16 videos
27 links
Сайт: https://exine.ai/

Приложение: @exine

Exine AI — трендовые нейросети без VPN в одном приложении. Единый баланс, пополнение картой РФ. Nano Banana бесплатно!
Download Telegram
🔥 Seedance 2.0 в Exine AI

Если нужен по-настоящему реалистичный результат и точное следование референсам и промпту, это одна из лучших моделей для видео на данный момент.

👍 Открываем Exine AI ➡️ Выбираем модель для создания фото ➡️ Nano Banana ➡️ Отправляем промпт:

Character sheet, same person in 5 views on pure white background (#ffffff): large full-body front view on the left; top center waist-up front; top right waist-up back; bottom center waist-up left profile; bottom right waist-up right profile. Photorealistic, studio lighting, ultra-sharp, high-detail, no shadows or reflections. Absolute consistency in face, hair, body, clothes, accessories, and shoes across all views


На выходе получаем развернутую карту персонажа — коллаж из 5 фото с разных ракурсов.

Далее переходим в Seedance 2, добавляем все референсные изображения и отправляем промпт:

## CORE SUBJECT
@image_1 = strict character reference: keep identical face, hair, beard, body proportions, skin tone, and identity.
@image_2 = strict outfit reference: keep identical clothing, colors, fit, and style.
@image_3 = strict skateboard reference: keep identical skateboard design and details.

Young male skateboarder, exact identity from @image_1, exact outfit from @image_2, exact skateboard from @image_3, filmed entirely in a front-facing selfie view. Full consistency for the entire 8s shot.
## ACTION
He holds a selfie camera, throws the skateboard onto the sidewalk, jumps on, and rides fast along a wide city sidewalk. Birds already on the sidewalk ahead scatter upward naturally as he passes, flying very close to the lens and around him. At an intersection, a car crosses from the side; he performs a realistic skateboard jump cleanly over the moving car and road space, lands, and keeps riding. Then he snaps his fingers: a visible cinematic wave appears and the whole world freezes instantly. Birds, people, cars, and all motion stop completely. Only he keeps moving. He rolls forward slightly, stops, pops the skateboard up, catches it, and walks forward through the frozen world.

## ENVIRONMENT
Wide modern metropolis avenue, broad sidewalks, road beside him, intersection, moving cars, pedestrians, bright daylight. He rides on the sidewalk, not in the car lane. Before the snap, the city is alive with motion. Birds begin naturally on the sidewalk and scatter through foreground and midground as he approaches. After the snap, everything is perfectly still: cars frozen, pedestrians frozen, birds frozen in mid-air close around him, only the main character moves.

## CAMERA / STYLE
Single continuous front-facing handheld selfie shot, 8s, no cuts, no angle changes, no external camera, no third-person view. Extreme wide-angle / GoPro feel, natural handheld shake, immersive forward motion. Cinematic realism, live-action look, realistic skin, fabric, skateboard physics, bird motion, daylight, HDR, urban detail, motion blur during speed, strong depth, no AI look.

## ACTION
The skater, matching @image_1, wearing @image_2 and riding @image_3, throws the board down, jumps on, and rides fast along the sidewalk of a wide modern city avenue. Birds already on the sidewalk burst upward and fly extremely close to the lens and around him. At an intersection, a car crosses from the side; he performs a realistic skateboard jump cleanly over the road space and moving car, lands, and keeps riding. Then he snaps his fingers: a visible cinematic pulse expands and the whole world freezes instantly. Birds, cars, pedestrians, and all motion stop completely. Only he keeps moving. He rolls a little farther, stops, pops the skateboard up, catches it, and walks forward through the frozen city.

## TIMING
0:00–0:02 throws board, jumps on, accelerates.
0:02–0:04 birds scatter close to lens; jumps over crossing car.
0:04–0:06 lands, rides, finger snap, instant time freeze.
0:06–0:08 rolls slightly, stops, catches board, walks.


🔗Полный гайд — оставили по ссылке

Переходи в Exine AI и тестируй Seedance 2 на своих идеях!
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
8👍2🔥2🤮1
Хочешь, чтобы визуалы из нейросетей выглядели сильнее?

👍 Мы собрали промпты для самых частых задач в GPT Image 2 и Nano Banana:

➡️ Карта персонажа
Create a clean professional character sheet / turnaround sheet in the same style and layout as the reference image, but ignore the black dots overlay completely. Do not include any markers, tracking points, dotted patterns, text, labels, props, or decorative elements.

Final image: studio-style character presentation arranged in 6 cells on one canvas, with a clean white or light gray background, soft even lighting, sharp focus, and realistic studio photography.

Layout:
Top row — 3 full-body views:
1. front view
2. back view
3. right three-quarter view

Bottom row — 3 chest-up / portrait views:
4. front view
5. left side profile
6. right three-quarter portrait

Requirements:
consistent character appearance across all views
consistent clothing across all views
accurate face and body proportions
clean spacing between cells
clear readable angles
high detail
professional model-sheet / turnaround-sheet aesthetic

Style: realistic studio photography, clean character sheet, minimal layout, professional fashion reference board.


➡️ Заменить одежду
Replace the clothes on the character from Image 1 with the outfit worn by the character in Image 2. Keep everything else from Image 1 exactly the same: character identity, face, body, pose, location, background, lighting, framing, and realism. Change only the clothing, and make it fit naturally and correctly.


➡️ Улучшить детализацию
Enhance this close-up portrait to a very high level of photorealistic detail while fully preserving the original person, identity, facial features, expression, framing, camera angle, lighting logic, and overall look.

Make the skin highly realistic with visible pores, fine skin grain, natural microtexture, subtle imperfections, realistic under-eye detail, lip texture, eyelashes, eyebrows, hair strands, baby hairs, sharp iris detail, and natural nose texture.

Do not beautify, smooth the skin, add glamour retouching, change the face, facial structure, makeup, expression, or make it look plastic, CGI, painted, or unnaturally over-sharpened.

Final result: an ultra high-resolution professional close-up with realistic skin detail, natural texture, and maximum photorealistic clarity.


➡️ Заменить локацию
Use @image1 as the main source image. Use @image2 only as the location/background reference. Replace the location in @image1 with the location from @image2. Very important: fully preserve the character from @image1 preserve the exact person, identity, face, hairstyle, body proportions, clothing, and all character details preserve the exact pose preserve the exact body position in frame preserve the exact camera angle preserve the exact framing and composition of @image1 preserve the exact scale of the character preserve the exact perspective of the character keep the character in the same position as in @image1 Change only the environment: place the character from @image1 into the location from @image2 adapt the new background naturally around the character make the integration realistic and seamless Important: do not change the character do not restyle the character do not change the pose do not change the expression do not change the clothing do not change the character’s placement in the frame do not add text Final result: the same exact character from @image1, in the same pose and same position, but inside the location from @image2. Format 16:9, 4k


🔗 Сохраняй и тестируй в Exine AI
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
17👍4🎉3🏆3🤝2👎1🤔1
Когда на фото нравится все кроме фона, его легко заменить

— без замены лица
— потери похожести
— и ручной возни

👍 Откройте Exine AI выберите раздел «Фото» → запустите Nano Banana, GPT Image 2 или другую модель для фото → загрузите изображение и вставьте промпт:

Используй первое фото как основу. Замени только фон и окружение, не изменяя самого человека. Полностью сохрани лицо, черты лица, прическу, мимику, позу, одежду, пропорции тела и общую узнаваемость человека. Новый фон должен выглядеть реалистично, с естественным светом, тенями и перспективой, чтобы человек органично вписывался в сцену.


Так можно быстро перенести фото в студию, город, лифт, офис, на съемочную площадку или любую другую локацию.

🔗 Приложение Exine AI
🔗 WEB Exine AI
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
9👍6🔥3🥰2👎1
This media is not supported in your browser
VIEW IN TELEGRAM
Создаём AI-мультфильм с любимыми героями с помощью Seedance 2 🎬

Как повторить:

1️⃣ Для начала выбираем фото героев, например волка и зайца из "Ну, погоди"

2️⃣ Открываем Exine AI нейросети для видео Seedance 2

3️⃣ Добавляем фото героев и промпт

Пример промпта:

SCENE 1 — CHILDHOOD (10 seconds)

PIXAR-style cinematic anthropomorphic wolf and rabbit toddlers sitting together in a noisy Soviet kindergarten cafeteria, both wearing cute pastel baby hats with little ears and pacifiers hanging on ribbons around their necks, long wooden cafeteria tables, aluminum bowls with oatmeal porridge, warm nostalgic Soviet kindergarten atmosphere, soft morning sunlight through windows, realistic fur details, emotional facial expressions

IMPORTANT: no music

SHOT 1 (0–3 sec)
Close-up: rabbit calmly eating porridge while wolf slowly turns toward rabbit with a sinister smug grin, looking directly into camera

💬 Волк, хриплый наглый голос:
— Зяяяяц…

SHOT 2 (3–6 sec)
Mid shot: wolf aggressively pushes metal bowl of oatmeal with elbow, porridge splashes all over rabbit’s face and long ears, children laughing loudly in background

💬 Дети:
— АХАХАХА!!

SHOT 3 (6–10 sec)
Close-up: rabbit sitting silently covered in porridge with sad lowered eyes while wolf laughs beside him with cruel smile

DreamWorks / Pixar cinematic realism, stylized realism, emotional storytelling, 8K

Negative prompt: gore, blood, horror, scary face, extra limbs, distorted anatomy, blurry image, ugly face


Для каждой сцены отдельная генерация — 🔗все промпты здесь

Переходи и тестируй свои идеи с Exine AI

🔗 Приложение Exine AI
🔗 WEB Exine AI
Please open Telegram to view this post
VIEW IN TELEGRAM
11🏆9🤝5👍2👎1🔥1
ИИ может сохранить не просто фото, а детскую фантазию. ☀️

Те самые рисунки, в которых для ребёнка спрятан целый мир: необычные существа, странные формы, яркие линии и придуманные герои.

В Exine AI все это можно превратить в реалистичную анимацию:

откройте приложение
выберите Seedance 2
загрузите детский рисунок

Промпт:
Transform this child’s drawing into a realistic, photorealistic scene. Keep the original composition, shape, proportions, and main colors, but make everything look real.
Turn the main subject into a believable real object or creature with natural textures, depth, lighting, and detail. Reinterpret abstract lines as real moving elements.
Add gentle natural motion, slight swaying, and a slow camera push-in. Cinematic realism, natural lighting, high detail. Duration: 5–6 seconds.
No crayon texture, no paper texture, no cartoon style. Make it look real.


🔗 Приложение Exine AI
🔗 WEB Exine AI
19👍8🍓2👎1🔥1🤮1💩1🤡1
Хочешь сделать ролик, где ты становишься персонажем игры? 🤔

Повторить можно в 3 шага:

1️⃣ - Открываем Exine AI → нейросети для фото → загружаем свою фотографию в GPT Image 2

Просим нейросеть превратить фото в карту персонажа в стиле нужной игры.

Важно: лицо, причёска и основные черты должны сохраниться, но визуал адаптируется под игровой стиль.

2️⃣ - Готовим локацию

Берём референс места, куда хотим попасть: город, арену, комнату, улицу, остров, замок или любую другую сцену.

Загружаем локацию в GPT Image 2 и просим адаптировать её под тот же стиль, что и персонаж.

Так персонаж и фон будут выглядеть как часть одного мира.

3️⃣ - Открываем Exine AI → нейросети для видео → Seedance 2

- Добавляем готовые изображения

- Дальше вставляем промпт для видео.

Промпт можно написать самому или попросить ChatGPT собрать его под конкретную игру, персонажа и сцену.

Для каждой сцены лучше делать отдельную генерацию — примеры промтов и статики под Sims

Переходи и тестируй свои идеи с Exine AI

🔗 Приложение Exine AI
🔗 WEB Exine AI
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
7🔥7👍3👎1🎃1