Vision Flow - News | Promts
3.35K subscribers
351 photos
130 videos
101 links
Подборка лучших промтов — вдохновляйся и создавай видео, музыку и изображения в нашем боте!
Наш бот с нейросетями - @vision_flow_bot
Download Telegram
1. Заходим в наш бот @vision_flow_bot
2. Выбираем: GPT2
3. Добавляем свое фото и вставляем этот промпт:

Grainy VHS camcorder footage from an American high school in February 1994. Random candid scene inside a school: hallway, cafeteria, classroom, gym, library, or parking lot. Low-resolution analog tape quality, washed-out colors, soft focus, scan lines, motion blur, fluorescent lighting, imperfect handheld framing, students in 90s outfits like flannels, varsity jackets, baggy jeans, backpacks, sneakers, and baseball caps. Make it feel like a forgotten home video frame from a school archive. Add a subtle timestamp with a random February 1996 date and time. No modern clothes, no phones, no HD, no cinematic look. You can also use this on your existing photos. Just upload your image first, then add this: "Transform this image into the same VHS high school camcorder style, while keeping the main person/ subject recognizable."


Vision Flow | News & Promts
1
1. Заходим в наш бот @vision_flow_bot
2. Выбираем: Nano Banana Pro либо другую нейросеть
3. Добавляем свое фото и вставляем этот промпт:

Create a high-definition, minimalist character design board on a white background, resembling a professional game art or fashion reference sheet. The layout should be neat with distinct sections, showcasing realistic and premium quality with consistent lighting and character features.

Left Side:
• Full-body three-view turnaround (front, side, back) of the same character with identical features, in a natural standing pose, suitable for modeling. Use eye-level camera angle and neutral lighting.

Right Side:
• Upper-right: Six headshot angles (front, top, back, side, near-side, 3/4 profile) showing consistent facial features and hair.
• Lower-right: Six close-up details (upper garment texture, lower-body clothing, hip tailoring, leg/skin texture, eyes/facial features, shoes).

Ensure all details match the character's appearance and are realistic, suitable for modeling references. The overall style should be minimalist, professional, and clean, with accurate material rendering, generous white space, and a refined finish.


4. Выбираем: Seedance 2.0
Загружаем фото из бананы, исправляем в промпте имя на своё и ожидаем результат.

Промт:
Style: Ultra-realistic Hollywood Oscars ceremony. Single continuous shot. Broadcast television camera mixed with handheld audience POV. Documentary realism. Audio: Live ceremony ambience — orchestral music, presenter announcing “And the Oscar goes to… НИНА!”, thunderous applause erupting, audience gasping, emotional crowd reaction, camera shutters firing.
Lighting: Grand theater interior. Warm golden stage lighting. Spotlights tracking the winner. Dark elegant auditorium with thousands of seated guests. Dramatic contrast between dark audience and blazing stage.
Setting: Dolby Theatre Hollywood. Rows of A-list celebrity audience in formal wear. Grand stage with iconic Oscar backdrop. Presenter standing at podium holding the golden statuette.
MAIN CHARACTER STRICT LOCK: Use uploaded face image and appearance, fabric catching stage light beautifully.
SCENE TIMELINE — 15 SECONDS
0–2s: Wide shot of grand auditorium. Presenter at podium opens envelope. Thousands of guests in anticipation. Spotlights on stage.
2–4s: Presenter leans into mic: “And the Oscar goes to… НИНА!” Thunderous applause explodes. Audience rises.
4–7s: Camera cuts to НИНА seated in audience. She places hand over mouth in shock, tears forming, rises from her seat. Guests around her applaud, embrace her.
7–10s: She walks the aisle toward stage, camera following her. Applause building.
10–13s: She ascends the steps, presenter hands her the golden Oscar statuette. She holds it up — golden statue gleaming under spotlights.
13–15s: She faces the audience, statuette in both hands raised, tears of joy, radiant smile. Full standing ovation. Camera flashes strobe across her face.
2
🎬Добавили новую Google Omni в бот!

Это полноценная «омни-модель», объединившая под одним капотом технологии Nano Banana, Genie и Veo для бесшовной работы с любыми типами данных.

Что умеет Gemini Omni:
🤩Omni способна симулировать физический мир, понимая законы гравитации и кинетическую энергию;

🤩Идеальная стабильность: Никаких «плывущих» кадров и артефактов! Omni железно удерживает консистентность персонажей и окружения при любых изменениях

🤩Первая и самая шустрая модель линейки — Gemini Omni Flash — уже интегрирована в бота и готова к вашим безумным идеям, а следом ожидаем релиз мощнейшей Pro-версии.


🎁 БОНУС В ЧЕСТЬ ОБНОВЛЕНИЯ!
Дарим всем промокод:
OMNI

Скидка 10%
Действует до 07.06

Vision Flow | News & Promts
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
1
Создаем свою фигурку LEGO

1. Заходим в наш бот @vision_flow_bot
2. Выбираем: Nano Banana Pro либо другую нейросеть
3. Добавляем свое фото и вставляем этот промпт:

Используйте загруженное пользователем изображение в качестве единственной ссылки на предмет. Превратите человека на загруженном изображении в реалистичную минифигурку в стиле LEGO, сохранив при этом его узнаваемую индивидуальность и одежду. Максимально сохраняйте сходство с лицом объекта в рамках дизайна минифигурки LEGO, сохраняя ту же прическу и цвет волос, адаптированные к шиньону LEGO, выражение лица, цвета одежды, дизайн и узоры одежды, а также любые видимые аксессуары. Результат должен четко представлять одного и того же человека, переведенного в форму LEGO. Персонаж должен соответствовать пропорциям минифигурки LEGO, включая цилиндрическую голову в желтом или телесном стиле, простые черты лица, адаптированные к изображению, туловище в форме блока с напечатанными деталями одежды, стандартные руки-минифигурки LEGO с изогнутыми кистями, короткие ноги-минифигурки LEGO и специальный формованный пластиковый парик, соответствующий изображению.ее прическа. Нарисуйте фигурку, используя реалистичный пластик LEGO с легким глянцевым блеском, тонкими формованными швами и естественными отблесками на поверхности игрушки. На лице и туловище должны быть нанесены печатные детали, характерные для минифигурок LEGO, а шиньон для волос должен быть выполнен из литого пластика. Конечное изображение должно представлять собой высококачественный реалистичный 3D-рендеринг игрушки с мягким рассеянным студийным освещением, тонкими тенями, подчеркивающими форму и текстуру пластика, и четким фокусом с четкими деталями. На кадре должно быть изображено все тело минифигурки с головы до ног в центральной композиции, при этом вся фигура должна быть видна полностью. Используйте простой белый фон без декораций, реквизита и дополнительных элементов. Окончательный дизайн должен быть похож на фотореалистичную игрушечную минифигурку LEGO с очень детализированной текстурой пластика и студийным освещением
Гуляем по Таймс-скверу в Нью-Йорке 😍 с помощью Gemini Omni 🔥

1. Заходим в наш бот @vision_flow_bot
2. Выбираем: Nano Banana Pro либо другую нейросеть
3. Добавляем свое фото и вставляем этот промпт:

Ultra-realistic minimalist character reference sheet on a pure white background, styled like a professional fashion photography and movie costume reference board. Left side: full-body front, side, and back standing views of the exact same person. Natural pose, arms relaxed, identical face, hairstyle, clothing, body proportions, and height. Eye-level camera, soft studio lighting, no background distractions.

Right side divided into two sections:
Top-right: 6 realistic head reference shots of the same person — front portrait, top/down angle, back of head, side profile, near-side angle, and 3/4 portrait.
Bottom-right: 6 close-up detail shots — clothing fabric texture, lower outfit detail, tailoring detail, skin texture, eyes/facial details, and shoes.

Style: ultra-realistic photography, premium fashion editorial quality, clean layout, sharp details, natural skin texture, realistic hair strands, accurate clothing materials, professional character identity sheet.


4. Выбираем: Gemini Omni
Качество: 1080 р
Соотношение сторон: 16:9
Длительность: 10 секунд

Загружаем свое фото (которое делали выше) и вставляем этот промпт:

[FACE REFERENCE: UPLOADED IMAGE]

GTA V native in-game graphics style. Use the uploaded photo as the exact facial reference. Preserve the person's recognizable identity, facial structure, hairstyle, skin tone, facial proportions, and key facial features while transforming them into an authentic GTA V playable character. The character must clearly look like the person from the uploaded image, but rendered in Rockstar Games' original GTA V character style.

Third-person gameplay camera. Full GTA V HUD visible.

Top-left mission box:
"Mission: Walk Through Times Square."

Top-right health, armor, and cash display:
$00015248

Bottom-left minimap:
"USA · New York · Times Square"

The player character based on the uploaded photo is walking through Times Square at night. Relaxed walking animation, natural arm movement, occasional head turns, observing the city surroundings, realistic GTA V character behavior.

Authentic Times Square environment recreated in GTA V style: giant LED billboards, Broadway signs, yellow taxis, NYPD cars, crowded sidewalks, tourists, food carts, digital advertisements, illuminated skyscrapers, busy intersections, reflective wet pavement, vibrant New York nightlife.

Dense NPC crowds. Tourists taking photos and videos. Street performers. Moving traffic. Dynamic city lighting. Steam rising from street vents. Realistic GTA V open-world atmosphere.

Camera follows behind the character exactly like standard GTA V gameplay. Smooth player-controlled movement. Slight camera sway. Seamless one-take gameplay recording.

Rendering style: original GTA V graphics engine, Rockstar Games lighting, GTA V shaders, authentic game textures, HDR lighting, realistic reflections, subtle motion blur, cinematic gameplay capture, ultra detailed 4K, unmistakably GTA V, not photorealistic, not cinematic AI movie style.
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
1
This media is not supported in your browser
VIEW IN TELEGRAM
Выпускной

1. Заходим в наш бот @vision_flow_bot
2. Выбираем: Gemini Omni
3. Загружаем две свои фотографии (Текущее фото и детское фото) Потом пишем такой промпт:

[Format: 16:9 horizontal, EXACTLY 10 seconds, ultra cinematic emotional short film]
IMPORTANT:
Use the uploaded reference photo as BOTH versions of the same person:
— younger child version (1st grade)
— older graduate version (11th grade)
The AI must preserve:
same facial identity,
same skin details,
same eyes,
same hairstyle style,
same ethnicity,
same recognizable appearance.
This prompt must work for BOTH boys and girls.
Background:
pure white seamless studio background,
minimalistic cinematic space,
soft natural studio lighting,
luxury emotional movie atmosphere.

0:00–0:04

Wide cinematic shot.
On the LEFT:
the younger 1st-grade version of the character, around 7 years old,
wearing classic school clothes,
holding white balloon:
“1 Класс 2016”
On the RIGHT:
the older 11th-grade graduate version,
wearing elegant graduation school outfit,
holding white balloon:
“Выпуск 2026”
They stand facing each other silently.
Very subtle slow camera push-in.
Natural breathing and blinking.
Emotional silence.

0:04–0:07

The older version slowly walks toward the child.
Without breaking eye contact,
the older version gently sits down on their knees in front of the younger self to become eye-level.
The child becomes emotional and slightly smiles.

0:07–0:10
The older version softly hugs the younger self.
At the same moment:
slow cinematic zoom into close-up.
Close-up emotional frame:
the child peacefully closes their eyes during the hug.
Soft warm lighting,
subtle cinematic grain,
realistic skin texture,
beautiful depth of field,
luxury movie look,
heartfelt emotional atmosphere about meeting your younger self.
2
This media is not supported in your browser
VIEW IN TELEGRAM
1. Заходим в наш бот @vision_flow_bot
2. Выбираем: Gemini Omni
3. Добавляем фото сетку, если есть детское фото, то тоже добавляем. А если нет, то только фото текущего времени, ИИ максимально сделает похожим.

Вставляем этот промпт:
[Format: 9:16 vertical, EXACTLY 10 seconds, ultra cinematic emotional video, NO voiceover, NO subtitles, NO talking]
Use the uploaded reference photo as BOTH versions of the same girl:
— younger first-grade version
— older 11th-grade graduate version
The AI must preserve the same facial identity, eyes, hairstyle vibe, ethnicity, and recognizable appearance.
Kazakhstan graduation atmosphere only.
Pure emotional storytelling through visuals only.

0:00–0:03
Wide cinematic shot.
A small first-grade girl slowly walks through a school graduation ceremony while holding hands with her future self.
The little girl wears:
white blouse,
large white bows,
dark skirt,
white tights,
classic first-grade school look.
The older version wears:
modern black school sarafan,
white elegant blouse,
graduation ribbon:
«Выпускник 2026»
Golden sunset light,
soft wind,
balloons,
flowers,
students blurred in background.
Very emotional atmosphere.
Smooth cinematic side tracking shot.


0:03–0:06
Medium emotional shot.
The little girl looks up at the graduate with admiration and safety.
The graduate looks down emotionally and softly smiles, almost holding back tears.
Natural emotional acting.
No exaggerated movements.


0:06–0:08
Front cinematic shot.
They continue walking hand-in-hand together through the graduation ceremony.
The older version gently squeezes the child’s hand reassuringly.
Warm sunlight wraps around them.
Beautiful shallow depth of field.


0:08–0:10
Slow cinematic close-up zoom.
Focus on their hands holding each other while walking toward bright sunlight.
Final emotional frame:
the child slightly leans closer to her future self while continuing to walk.
Ultra realistic skin texture,
Netflix-style cinematic lighting,
soft film grain,
luxury emotional movie atmosphere.
Генерируем шарики со своим лицом с разными эмоциями

A surreal and humorous wide-angle photograph of [REFERENCE SUBJECT] standing exactly as in the reference image. The subject is wearing the same clothing, colors, pose, and proportions as in the reference.
Surreal Elements: Floating above the subject, like a bouquet of balloons, are five oversized, hyper-realistic, disembodied heads of the same subject. Each head has the same facial features and hair as the subject and is making a different, highly expressive, funny face (e.g., exaggerated pout, shocked gasp, bored eye-roll, side-eye gaze, slight smirk).
Details & Style: The subject holds a cluster of five delicate, hand-drawn-style white strings connecting to the floating heads. The background remains exactly as in the reference image, with all lighting, objects, and environment preserved. Bright, direct sunlight (or the same lighting as reference) casts realistic shadows. The aesthetic blends avant-garde fashion photography, realism, and whimsical digital art. Ultra-high definition, 8k resolution, photorealistic textures, cinematic lighting.
Optional Notes: Ensure the floating heads are surreal and humorous, but the subject and the background remain identical to the reference image. Maintain proportions, pose, and realistic textures.