深夜便利店的门口,雨后街道还泛着霓虹微光。她拎着一杯冰咖啡走出来,路人视角随手一拍——又美又真,这条提示词值得复制。
Today's prompt 👇 (copy-paste ready):
Ultra-realistic 10-second vertical 9:16 candid street video, night. A beautiful young East Asian woman, early 20s, fresh natural beauty, realistic skin with slight sheen, minimal makeup, shoulder-length dark hair with natural volume. She wears a simple beige trench coat over a black turtleneck, holding an iced coffee. Same face, hairstyle and clothing throughout.
Outside a convenience store at night: cool fluorescent spill from the entrance mixing with warm street-lamp glow, wet pavement reflections after light rain. Filmed handheld on a phone like a passerby, uneven framing, mild digital noise, autofocus hunting, no beauty filter.
0:00–0:05: She steps out of the store, iced coffee in hand, glancing down the street, neon signs reflecting softly on the wet ground.
0:05–0:10: She notices the camera, gives a brief natural smile and a small nod, then continues walking, coat swaying gently.
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #PromptShare #夜景
AI 短剧成本降了 90%,Seedance 2.5 却反而涨价约 50%——工作室们现在玩的是「分层用工具」:关键镜头(人物特写、情绪转折)交给 Seedance,要求不高的镜头(转场、配角走位)分流给可灵、万相等便宜工具;预算实在吃紧,还有第三方代理平台能做到官方价的 1-4 折。好钢用在刀刃上——钱包和画面都保住 💰
Video prompt (copy-ready):
A dignified ancient Chinese beauty in an elegant pale-cream hanfu with wide flowing sleeves, delicate floral hairpins and pearl earrings, walking slowly through a spring courtyard filled with blooming pink cherry blossoms; red lanterns hang from a wooden pavilion nearby; soft sunlight filters through petals drifting in the breeze; camera slowly pushes in; photorealistic portrait, cinematic lighting, ultra-detailed fabric texture, serene and graceful mood, vertical 9:16, 10 seconds.
🎬 本视频由 AI 生成 | AI-generated video
#AI短剧 #工具横评 #Seedance #PromptShare
晨光里的办公室,拿铁还冒着热气 ☕ 写实风 OL,美得不像 AI 画的。
Ultra-realistic cinematic video of a stunningly beautiful young Asian woman in her mid 20s, delicate refined features, luminous natural skin with subtle realistic texture, soft elegant makeup, long silky black hair. Tailored white blouse and elegant knee-length pencil skirt. Chic modern office lounge, golden morning light through floor-to-ceiling windows. She smiles softly at the camera holding a latte, hair swaying lightly. Shallow depth of field, gentle handheld movement, photorealistic, 8K, DSLR, 85mm lens.
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #AIArt #PromptShare
☁️ 仙侠云海 · 白衣御风。低空 FPV 掠过云海仙山,衣袂翻飞,一镜飞越人间仙境。
Prompt to copy:
"Ancient Chinese beauty in flowing white hanfu flying horizontally at low altitude, high-speed FPV drone perspective skimming over a sea of clouds, misty immortal mountains, deep canyons and red temple architecture, arms swept back, long black hair and wide sleeves trailing in the wind, dynamic motion blur, photorealistic portrait, cinematic lighting, vertical 9:16"
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #AIArt #PromptShare
同题材开工前,先花 30 秒看这张横评 👇
古风美女做视频,起手方式不同,成品差距巨大——图生视频 VS 纯文生视频,我的实测结论:
📍 图生视频(供人设图起手)
✅ 人脸、服装、发型锁死,系列化运营必备
✅ 动作和运镜只用描述场景,不用再"抽卡"赌长相
⚠️ 人物动起来后偶尔走样,每条都要逐帧自审;单条约 10 秒
📍 纯文生视频(只写 prompt)
✅ 创意自由:任何动作、机位、场景一句话起手
✅ 适合做概念测试、快速验证题材方向
⚠️ 同一人物无法复刻,不适合做固定人设账号
📌 选型建议
做账号、做系列 → 图生视频起手,先定人设图再谈动作
找灵感、试题材 → 纯文生快速出样片,跑通后再转图生定版
评论区说说你平时用哪一路 👇
英文 prompt 示例(可复制,图生视频飞行题材实测写法):
[your character image] + An ancient Chinese beauty in flowing white hanfu flying horizontally over a sea of clouds, arms extended backward, sleeves billowing in the wind, long hair streaming; fast low-altitude FPV tracking shot following closely, soaring over misty mountains and valleys, replace any studio background with a real cloudscape, slow push-in, vertical 9:16, about 10 seconds, photorealistic, cinematic lighting
#AIVideo #PromptShare #AIArt
山风拂过剑穗,白衣剑客立于云雾山巅——一剑挥出,剑气如虹。⚔️
Photorealistic portrait of a dignified ancient Chinese swordswoman in flowing white hanfu, standing atop a misty mountain peak at dawn, slowly drawing her long sword and sweeping it in a graceful arc, wide sleeves and long hair billowing in the mountain wind, ancient pine trees and a drifting sea of clouds behind her, cinematic lighting, slow orbital camera push-in, vertical 9:16, elegant wuxia atmosphere, film-grade detail
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #AIArt #PromptShare
@momochan404 Yes, exactly! I keep skin tone on its own line — something like "natural, healthy warm skin tone, clean and true-to-life" — and let the neon live in a separate lighting line. That way the model colors the scene with neon but doesn't tint the skin pink or green. 🌃
1930年代老上海外滩,夜色、霓虹、旗袍、老爷车 🌃
她回眸一笑,转身侧身坐进车里,轻轻挥手告别。
Grok Imagine 一张图 + 一段动作提示词,15秒一镜到底,提示词直接抄👇
【① 图片提示词】
A glamorous 1930s Old Shanghai beauty standing beside a black vintage 1930s limousine on the Bund at night, sleek dark hair in elegant finger-wave curls with a jeweled hairpin, porcelain skin, arched brows, soft smoky eyes, deep red lipstick, pearl drop earrings, long pearl necklace, a fitted high-collar crimson silk qipao with gold floral embroidery and a high side slit, sheer black gloves, a white fur stole over one arm, holding a small beaded clutch, looking at camera with a calm confident half-smile. Behind her: the classic Bund skyline with the Customs House clock tower and neoclassical stone buildings lit in warm gold, vintage neon signs in Chinese characters glowing red and green, a rickshaw passing in soft blur, light mist over the Huangpu River. Wet cobblestone street reflecting golden and red neon, chrome headlights catching the light, cinematic night photography, 1930s film-noir glamour lighting, shallow depth of field, subtle film grain, photorealistic 8k --ar
【① 动作提示词(15秒)】
Use the image as the first frame and keep her face, hairstyle, qipao and the Bund background exactly the same. 8-second cinematic shot, slow push-in from medium to close-up. She rests one gloved hand on the car door, slowly turns her head toward the camera, then lowers her eyes shyly for a moment and tucks a loose curl behind her ear. She lifts her gaze back to the lens with a soft knowing smile and gently pulls the fur stole higher on her shoulder. The qipao silk shimmers as she shifts her weight, the pearl necklace sways slightly. In the background, neon signs flicker, the rickshaw passes by, mist drifts over the river, and puddles ripple with reflected light. Smooth natural motion, stable identity, no face distortion, no extra limbs, no text, no subtitles.
【② 动作提示词(15秒)】
Use the image as the first frame and keep her face, hairstyle, crimson qipao, pearls, fur stole and the Bund night background exactly the same. 15-second continuous cinematic shot, no cuts.
0–4s: She stands beside the black 1930s limousine, slowly turns her head toward the camera, lowers her eyes shyly for a moment, then tucks a loose curl behind her ear.
4–7s: A uniformed chauffeur's gloved hand opens the rear door (only his hand and sleeve visible). She gathers the fur stole on her arm and slowly turns around, her back now facing the camera, the qipao hugging her figure, pearls swaying at her neck.
7–11s: With her back to the camera, she turns her body sideways to the open rear door, lifts the hem of her qipao slightly with one hand, bends gracefully and sits down sideways on the leather rear passenger seat, then swings both legs inside together in a ladylike motion. She sits in the rear passenger seat, not the front seat.
11–15s: Camera moves closer to the open window. She settles into the seat, glances back at the lens over her shoulder with a soft knowing smile, and gently raises her gloved hand in a small farewell as the door closes.
Background: neon signs flicker in red and green, mist drifts over the Huangpu River, a rickshaw passes, wet cobblestones ripple with golden reflections. Smooth natural motion, stable identity, no face distortion, no extra limbs, no text, no subtitles.
#Grok #GrokImagine #AI视频 #老上海 #旗袍
深夜特别版——9/26 最高浏览的"第一人称视角"复刻来啦:雪夜红衣,带你第一视角穿越峡谷,衣袂飘飘。
First-person FPV perspective, ancient Chinese beauty in crimson hanfu flying horizontally at low altitude through a snowy canyon at night, arms swept back, sleeves and long hair trailing, snow swirling, high-speed drone flight skimming canyon walls, photorealistic, cinematic lighting, vertical 9:16
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #PromptShare #AIArt
竹林深处,一曲抚琴。
今日提示词:竹林琴女——静谧系古风小品,微风拂动衣袖与长发,镜头缓缓环绕推进。🌙
Prompt for reference:
Dignified Chinese beauty in flowing hanfu standing in a bamboo forest pavilion, playing the guqin, sleeves and long hair swaying in the breeze, camera slowly orbiting closer, moonlight through bamboo leaves, photorealistic portrait, cinematic lighting, serene atmosphere, vertical 9:16
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #AIArt #PromptShare
御剑飞行也有公式:低空 FPV 视角给速度感,地形纵深做参照,水平平飞双臂后展给动感,白底强制替换成真实场景防穿帮,最后配上固定 BGM 收束情绪。这就是我们飞行模板的五层拆解,换题材直接套用——今天这条是仙侠云海版。
Prompt (copy & share):
"Ancient Chinese beauty in flowing white hanfu, flying horizontally at low altitude above a sea of clouds and mountain peaks, high-speed FPV drone perspective, arms swept back, long hair and wide sleeves trailing in the wind, dynamic motion blur, photorealistic portrait, cinematic lighting, vertical 9:16, graceful and dignified"
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #PromptShare #AIArt
千年前壁画里的飞天,今夜有了真实的影子。敦煌飞天系列第一弹:莫高窟前的月夜飞行,飘带漫天。
Photorealistic portrait of a graceful ancient Chinese woman in a flowing ivory Hanfu dress with wide sleeves, golden hairpins and jade earrings, flying elegantly like a Dunhuang flying apsara in front of the Mogao Caves at night. Ancient murals and Buddha statues behind, swirling mist, glowing red lanterns, long ribbons dancing in the wind, slow orbiting camera push-in, cinematic lighting, warm and cool color contrast, elegant ancient Chinese aesthetic, vertical 9:16 video.
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #AIArt #PromptShare #敦煌飞天
Yes, I do! For the neon glow I usually add a style block like "neon-noir cinematic lighting, warm neon signs reflecting on wet cobblestones, soft bloom but keep skin tones natural." A reference image also helps a lot — especially for the color temperature of vintage neon. Try it and share your result! ✨
烟雨江南,紫衣佳人执伞过桥。回眸一瞬间,石径青苔也温柔了起来 ☂️
Prompt for fans to copy:
An elegant photorealistic Chinese woman with an ornate hair bun adorned with purple flowers, wearing a long purple high-neck halter gown, holding a red floral oil-paper umbrella, walking slowly along a misty Jiangnan water-town alley after rain, stepping onto a mossy stone arch bridge over a canal, turning back to look at the camera with a gentle smile, camera slowly pushes in, willow trees and wooden boats in background, soft diffused natural light, cinematic lighting, photorealistic portrait, vertical 9:16, 10 seconds
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #PromptShare #AIArt
白衣仙子贴水飞行,江河峡谷极速掠过 🌊 完整提示词附上,拿去用👇
A graceful woman in a modest white hanfu dress flying at ultra-low altitude over rivers and canyons, body stretched horizontally, arms swept back, dress flowing in the wind, FPV drone perspective, extreme speed with cinematic motion blur, photorealistic portrait, cinematic lighting, 8K, vertical 9:16
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #AIArt #PromptShare
Thanks so much! 🙌 For the Old Shanghai neon vibe, I layered the prompt with period-specific descriptors: "1930s Shanghai street, vintage neon signs with Chinese characters glowing in warm red and teal, wet cobblestone reflections, Art Deco facades, rickshaw lamps." The trick is to name the era explicitly (1930s, Republic of China era) and separate lighting cues from architecture — glowing signage + wet-ground reflections really sell the period mood. Hope that helps! ✨
OpenAI 昨天发布了 textGrain 隐形水印——AI 生成的文本现在能被机器验出来了,加上 X 自己也会自动识别 AI 内容,想靠“混过去”越来越难。创作者保号最稳的一招反而最简单:每条都主动披露。本账号的双重披露规范拿去抄作业 👇
OpenAI just announced textGrain, an invisible watermarking system that lets AI-generated text be detected by machines (EU rollout underway via the EU AI Act). X also auto-detects AI content — verified on this very account. The safest play for creators is proactive dual disclosure: a disclosure line in the caption plus the platform's "Made by AI" toggle after posting. Details 👇
🎬 本视频由 AI 生成 | AI-generated video
#AI合规 #技巧干货 #AIGC
荷塘月色,古风佳人执伞临水,回眸一笑的氛围感拿捏住了🌙 喜欢这条就复制下面的英文提示词自己试试~
Prompt:
A stunning ancient-Chinese-style beauty, big eyes, high nose bridge, oval face, elegant figure, wearing a low-cut ancient costume with key areas modestly covered. Scene: moonlit lotus pond at night, a bright full moon hanging in the sky, lotus flowers and lily pads covering the water, fireflies drifting in the air. She holds a silk oil-paper umbrella, standing on a lakeside stone arch bridge, gentle breeze moving her ribbons and dress hem, she slowly turns back toward the camera and smiles. Camera: slow push-in from wide shot to half-body, soft moonlight on her and shimmering water, serene and poetic mood. Ancient Chinese aesthetic, cinematic quality, refined and dreamlike.
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #AIArt #PromptShare
竹林晨雾未散,琴声惊起一片竹叶。
这是今日份古风美女提示词——竹林琴女🎋 喜欢就拿去用!
A breathtaking ancient Chinese beauty with big bright eyes, a high nose bridge, a delicate oval face, and a full curvy figure, wearing an elegant low-cut traditional hanfu with key areas modestly covered. Setting: a misty bamboo forest at dawn, sunlight filtering through bamboo leaves. She sits in a bamboo pavilion playing a guqin; the music seems to stir a bamboo leaf to drift down. The camera slowly pushes in as she looks up and smiles softly. Vertical 9:16, about 10 seconds, photorealistic cinematic quality, classical Chinese aesthetics, smooth natural motion.
🎬 本视频由 AI 生成 | AI-generated video
#AIVideo #AIArt #PromptShare
8GB 显卡也能本地跑 MiniMax H3 了!FlashML-org 开源的 FreeVideo(Apache 2.0),把八步蒸馏版 VDN-H3 搬进消费级显卡,门槛降到 8G 显存 + 16G 内存。代价是生成慢、严重吃 SSD,官方 2K 能力也只在托管端。结论:想尝鲜练手 → 本地跑 FreeVideo;要交付质量 → 还是托管 API。
下面附本条视频的英文 prompt,同一条拿去 FreeVideo 和托管 API 各跑一遍,对比效果最直观 👇
Prompt to copy:
A stunning adult woman in an ornate Tang-dynasty palace gown, fitted silk ruqun with a flowing peibo scarf, fully clothed, elegant curves accentuated by tailored fabric, walking slowly along a golden palace corridor, silk gown swaying gently in the wind, camera slowly dollying in from a side wide shot to a half-body close-up, cinematic palace lighting, imperial Chinese aesthetics, photorealistic detail, 9:16 vertical, 10 seconds
🎬 本视频由 AI 生成 | AI-generated video
#MiniMaxH3 #AI视频 #工具横评 #开源