造梦CreateDream retweeted
Sd2.5参考视频
这几天一直在试sd2.5的人物替换功能,但是由于卡真人卡的厉害,只能借助参考视频来实现这样的逻辑。
由于sd2.5的优秀的模仿能力,基本能还原原视频的动作,而且相比Wan3最大15s参考视频,支持多个视频片段参考。
无审查AI平台:reurl.cc/nxby7D
🤖 Made with AI
造梦CreateDream retweeted
This prompt didn’t get picked at first, but looking back, the more I see it the more I like it 😆
Grok Imagine 2.0 :
Photorealistic editorial portrait.
The setting is close grassy ground plane, dense green lawn texture, short grass blades.
She has her chin slightly lowered, and presents a sitting pose.
Waist-up portrait, front view. Photographed directly from the front. Her chest, torso, and waist face the lens in a frontal waist-up silhouette.
A 20s seductive stunning Japanese or Korean woman with clear transparent frame, thin-frame glasses, small cross earring detail, subtle gothic-inspired accessory accent, and layered cross pendant necklace detail, subtle gothic-inspired neck accent, I–K cup bust, slim arms, narrow waist, wide hips, fuller upper thighs, slim calves, balanced proportions, pronounced hourglass silhouette, curves stretching fitted fabric with soft flesh bulges at snug garment edges; height 163–168 cm, bust–waist–hip 103–60–98 cm, young seductive alluring beauty face, deep side-parted long soft waves, defined face-framing layers, polished wave shape, lightly wind-swept movement, gentle directional flow, a few naturally loose strands, controlled overall silhouette, natural black hair, soft natural smile, relaxed brows, gently narrowed eyes, softly parted lips, lifted cheeks.
She wears fitted upper-body cut following the garment shape, off-white silk camisole top, delicate straps, fluid sheen, worn untucked, its hem draping naturally over the waistband and upper portion of the lower garment, partially concealing it while leaving the remaining fabric visible below, ultra-low-rise waistband sitting very low on the hips, tight body-skimming lower-body fit, black seamless lace thong bottoms, ultra-thin side straps, low-rise V-front, minimal back panel at the lower crop edge.
Yoshihiko Ueda-inspired quiet natural image language.
Fujifilm Superia film rendering, cool-green shadow tint, punchy reds and greens, medium-high contrast curve, visible consumer-film grain.
🤖 Made with AI
Grok Imagine 2.0 :
Photorealistic editorial portrait.
The setting is horizonless seamless matte pure white color field, continuous white ground-and-background plane blending into a solid white void, full-bleed white surface.
She has her chin slightly lowered, and presents a loose, heavy upper-body slouch with dropped shoulders and a relaxed torso.
Waist-up portrait, front view. At a relatively close portrait distance, roughly 1.5–2 meters from her, the camera is positioned above her and angled downward, revealing the top planes of her shoulders and waist. Photographed directly from the front. Her chest, torso, and waist face the lens in a frontal waist-up silhouette.
A 20s seductive stunning Japanese or Korean woman, I–K cup bust, slim arms, narrow waist, wide hips, fuller upper thighs, slim calves, balanced proportions, pronounced hourglass silhouette, curves stretching fitted fabric with soft flesh bulges at snug garment edges; height 163–168 cm, bust–waist–hip 103–60–98 cm, young seductive alluring beauty face, deep side-parted long soft waves, defined face-framing layers, polished wave shape, lightly wind-swept movement, gentle directional flow, a few naturally loose strands, controlled overall silhouette, natural black hair, soft natural smile, relaxed brows, gently narrowed eyes, softly parted lips, lifted cheeks.
She wears fitted upper-body cut following the garment shape, off-white silk camisole top, delicate straps, fluid sheen, worn untucked, its hem draping naturally over the waistband and upper portion of the lower garment, partially concealing it while leaving the remaining fabric visible below, ultra-low-rise waistband sitting very low on the hips, tight body-skimming lower-body fit, black seamless lace thong bottoms, ultra-thin side straps, low-rise V-front, minimal back panel at the lower crop edge.
Yoshihiko Ueda-inspired quiet natural image language.
Fujifilm Superia film rendering, cool-green shadow tint, punchy reds and greens, medium-high contrast curve, visible consumer-film grain.
造梦CreateDream retweeted
通用🔵 GPT image 2.5 Prompt:
16宫格姿势图×Editorial人像摄影×主体:【XXX】
生成1张完整的9:16竖版4×4人物姿势图,共16个独立画面。
最高优先级:16格身体结构严格对应01至16,不能仅通过镜头角度、人物朝向、表情或裁切制造差异;差异必须来自身体结构,每格使用不同的身体支撑、腿部或躯干结构。
一、主体一致性
16格保持同一主体的身份、外貌、体型、发型、发色、服装、配饰与核心视觉特征一致;头发、衣摆、褶皱可随动作自然变化。主体始终为第一视觉主体。主体身份、服装与视觉特征不得因姿势变化而漂移。
二、16种姿势
01后方跪坐式
双膝着地,臀部向后靠近脚跟,身体背向镜头,上身保持直立,形成明确的后方人物轮廓。
02侧展腿跪坐式
双膝着地,一条腿向身体侧方展开,另一条腿保持弯曲,上身直立,形成明显的侧向腿部线条。
03四足斜向跪姿式
双手与双膝着地,形成明确四点支撑,身体呈明显斜向结构,一侧腿向侧前方展开。
04俯卧单腿抬起式
身体完整俯卧,上身由前臂支撑,一条腿向后上方弯曲抬起,另一条腿保持伸展。
05纯侧卧伸腿式
身体完整侧卧,一侧手臂支撑上身,双腿沿身体方向伸展,形成清晰的纯侧面轮廓。
06侧坐后撑伸腿式
坐姿,一条腿向前伸展,另一条腿弯曲收拢,身体向后倾,一只手在身后撑地承担重量。
07正面展开跪坐式
双膝着地,身体正对镜头,双腿向左右两侧大幅展开,双手自然放置。
08单臂抬手坐姿式
坐姿,一条腿弯曲收拢,另一条腿自然展开,一只手向上延伸,另一只手支撑或保持身体平衡。
09俯卧前伸式
身体完整俯卧,双腿向后伸展,双臂向身体前方充分伸直,形成明显的纵深方向。
10后方侧跪伸腿式
身体背向或侧背向镜头,双膝着地,一条腿向身体侧后方充分伸展,上身保持直立。
11仰卧屈膝式
身体完整仰卧,背部贴地,双膝向上弯曲,双脚靠近地面,头部与肩部保持贴地。
12极低位后方四足式
双手与双脚着地,臀部明显抬高,身体形成紧凑的倒V形结构,呈高臀低身的四足结构,从后方低机位观察。
13坐姿环抱腿式
坐姿,双腿向身体前方收拢并交叠,双臂环抱双腿,上身略微向前收缩。
14侧卧托头屈腿式
身体完整侧卧,一只手托住头部,双腿同时弯曲并向身体方向收拢。
15仰卧抬腿式
身体完整仰卧,双腿明显抬离地面并向上弯曲,形成突出的纵向腿部结构。
16后撑侧坐收腿式
坐姿,双腿完全收拢并向身体一侧折叠,身体明显后倾,一只手在身后撑地,另一只手自然放置,形成紧凑的侧向轮廓。
16个姿势必须有明确身体结构差异,严格按01至16执行,不交换、不重复、不遗漏。
三、宫格构成
严格排列:
01 02 03 04
05 06 07 08
09 10 11 12
13 14 15 16
每格仅呈现一个完整主体和对应姿势。统一Editorial人像摄影语言、人物尺度、服装系统与背景逻辑。人物完整入镜,身体结构、手脚膝肘及支撑关系清晰。格间边界清晰,不重叠、不串格、不交换姿势、不遗漏、不重复。
四、背景系统
根据主体服装、发色、代表色与视觉身份提取主色及邻近色,最多两种核心色相,生成低饱和、低对比度极简渐变背景。
仅使用大面积纯净色彩与缓慢渐变,不添加环境、道具、纹理、复杂光影或装饰元素。
16格共享同一渐变逻辑,保持统一、干净、连续,并保留大面积负空间。背景不得形成具体场景。
五、16宫格隐藏视觉
将16个背景视为连续的大型渐变平面。
固定隐藏视觉:主体本身的低对比度45度侧脸轮廓。
根据主体脸部特征、发型与核心视觉特征生成轮廓,以极弱明暗变化融入背景,并连续拆分为16个片段,按4×4自然衔接。
隐藏轮廓仅使用当前主体特征,不引入其他人物;保持低对比、低饱和、弱可见,仅存在于背景负空间。
仅表现抽象连续的明暗形态,不呈现眼睛、鼻子、嘴巴等具体五官。
单格只能看到局部明暗形态,完整观看4×4时才能辨认完整侧脸。
不得覆盖主体、干扰姿势、添加额外人物、形成第二主体或产生水印效果;隐藏轮廓不得成为独立人脸。
六、视觉层级
第一层:主体与16种姿势。
第二层:极简低饱和渐变背景。
第三层:16格连续隐藏视觉。
主体清晰度优先于背景连续性,背景连续性优先于隐藏视觉。
七、最终输出
只生成1张9:16竖版完整图片,内部为4×4共16个独立人物画面。
严格执行01至16,不遗漏、不重复。
不添加标题、说明文字、Logo、杂志排版文字或其他版式文字。
单格为完整、清晰的Editorial人像姿势画面;整体形成统一的16宫格,并在背景中隐藏由16格共同组成的主体45度侧脸轮廓。
🤖 Made with AI
造梦CreateDream retweeted
MINIMAXH3提示工具已经更新184版本,更新了页面和镜头,人物动作表情控制更加的精准,欢迎体验
提示词工具: h3keys.com
视频都是造梦ai生成:@createdream001
#提示词 #成人提示词 #minimaxH3提示词 #提示词工具 #Grok提示词 #grok提示词工具 #GROK成人提示词
🤖 Made with AI
造梦CreateDream retweeted
《The Black Master》
人物测试样片
Sample 02
图片生成:chatgpt - gemini - grok - comfyui
图片转视频使用工具👇
视频AI:reurl.cc/8YOW0g
参考提示词:aivideo0.com/blog
案例:aivideo0.com
不要再问我咋个做,不怕烧钱你们就自己来试~~
All characters, scenes, and visuals in this video are entirely AI-generated from prompts and do not depict or involve any real person.
#亚洲熟女 #黑人 #媚黑 #偷情 #出轨 #AIFilm #NSFW #AIArt #MILF
🤖 Made with AI
造梦CreateDream retweeted
《Black Man × Mature Asian Woman》
Sample 01
Still considering whether to add a third person. The storyline is still being revised.
使用工具👇
视频AI:reurl.cc/8YOW0g
参考提示词:aivideo0.com/blog
案例:aivideo0.com
#亚洲熟女 #黑人 #媚黑 #偷情 #出轨 #AIFilm #NSFW #AIArt #MILF
🤖 Made with AI
造梦CreateDream retweeted
一觉醒来 GPT-Image 2.5 更新了。
第一件事:想到拿复杂的多镜头Prompt测故事版。👀
结果有点离谱:之前 2.0 容易漏掉的镜头,这次几乎全部还原出来了。而且速度还快了不少。
我一直都有个习惯,
就是先用故事版检查视频 Prompt。
这样可以有效减少积分浪费,
也能提前看看 Prompt 的思路和画面
🤖 Made with AI
造梦CreateDream retweeted
用GROK和造梦ai手搓小短剧,一定要看到最后,视频2K高清,你的点赞和转发是我创作最大动力 ,视频GROK和造梦ai生成
造梦ai一键生成无审查短剧:
reurl.cc/NOlKjk
一张图片即可生成劲爆的ai视频工具: reurl.cc/rkDrGb
reurl.cc/1lvk8X
造梦ai图片提示词:【第一人称视角,POV,我跪着,从我眼睛往下俯视,一个性感美女仰躺着,头朝图片上方,双腿张开,女人冷艳中带着媚意,柳眉凤眼,大乳房,细腰,她的胯部和臀部暴露,阴部湿润诱人。我的阴茎插进她湿润的阴道里,盘头,长发披肩,发饰,,表情色气。写真风格,真实摄影,电影镜头。】
动作提示词:
【男人把阴茎插件女人的阴道深处,她发出娇喘,动作流畅,电影镜头,不会出现不合理的画面,男人快速的耸动着,女人的乳房伴随着男人的耸动抖动,阴茎每次都插到阴道最深处。最后男人用手揉着女人的奶子,全程阴茎抽插阴道的动作不停】
图片提示词:
【第一人称视角,POV,我站着,从我眼睛往下俯视,我站在女人的后面,我胯部贴着女人的臀部,我双臂紧紧抓住她的臀部,阴茎插入她的阴道里。粗勃起的阴茎、女人趴在桌子上,头朝图片上方,回头看着我,,女人冷艳中带着媚意,柳眉凤眼,大乳房,细腰,她的胯部和臀部暴露,阴部湿润诱人。我的阴茎插进她湿润的阴道里,盘头,长发披肩,发饰,,表情色气。写真风格,真实摄影,电影镜头。】
动作提示词:
【影感的中景镜头他粗壮、勃起的阴茎以高速、有节奏的抽插动作在她肛门中进出。她丰满的乳房随着每一次撞击剧烈晃动,长长的头发前后摇摆。她的臀部随着男人骨盆重重撞击她的屁股而抖动并分开。女人的臀部随着每一次抽插抖动和弹跳,她的嘴张开,发出愉悦的呻吟和尖叫。场景激烈且充满动态,捕捉到摩擦和他们下半身剧烈的运动。随着场景的进行,男人加快节奏,抽插得更快、更用力,他的腹股沟撞击她的臀部,震动着女孩的身体。不会出现不合理的画面,女人保持姿势】
图片提示词:
【第一人称视角,POV,我站着,从我眼睛往下俯视,女人冷艳中带着媚意,柳眉凤眼,大乳房,细腰,,盘头,长发披肩,发饰,,表情色气。跪在我的面前全裸身体,写真风格,真实摄影,电影镜头。女人跪在男人面前。男人握着双手抱住她的头,女人张大嘴巴含住男人粗大的阴茎。】
动作提示词:
【女人跪在男人面前。男人握着她的头,将她的头前后移动,推动和拉动她的头靠近阴茎。她在深喉他的阴茎。阴茎反复进入她的嘴里。她的头前后摆动。她的脸上覆盖着精液,精液挂在她的下巴上。她抬头看着男人寻求帮助。男人将她的头按在阴茎底部一秒钟,保持她的头紧贴他的腹肌。她无法呼吸,呛咳着:“MMMMMMMMM!!!”。动作 jagged。她把手放在他的腿上,试图推开自己,但无法抗拒】
大家可以试试效果真的很好。
#角色扮演 #二次元 #少妇 #ai电影 #御姐 #火辣 #性感 #清純 #寫真 #古装美女 #短剧 #ai短剧 #巨乳 #母狗
🤖 Made with AI
造梦CreateDream retweeted
Grok 做成人视频如何写提示词全是干货,记得点赞
1.很多人不是不会写,是 额度先被自己点没了。失败、重抽、瞎开高清,一周就没了。下面大白话。
2.提示词四步:人物 → 机位 → 几秒做什么 → 禁止什么。别只写「做爱」。
3.机位优先 侧面。少用 正面大特写(又拒又糊又费次数)。
4.少被拒:少写器官;多写 贴紧、无脸、只露一点、四角裤。「后入」改 胯贴着屁股;秒出现用 两段硬切。
5. 人物一致:定妆图当锚,人设复制不改字,每条加「同一张脸、不换脸」。 画面一致:同一房间、光、衣服、机位;「场景不变,只改动作」。
6.第一人称写清 我能看见什么(自己肚子/手/四角裤 + 她后背),并写 看不见自己的脸。
7.人要大:写 占满画面、少留白。难姿势:先出对的静图再编辑,别整张重抽。
8.额度是 每周共享池。流程:静图 → Grok 只出 480 → 剪映拉 2K。不开 720。 提示词写好再点;被拒先改词;能编辑续作就别新建。
9. 例子|趴桌第一人称(克制)主观第一人称,站立男性微低头。自愿成年人。男穿四角裤。能看见:自己的腹部和手,女人上身伏桌、双腿站立的后背与腰。男人站她身后,腰胯轻轻贴着她,手扶腰。不见自己的脸。写实。 禁止:男人露脸,夸张动作,未成年。
10. 例子|跪地侧面(男人只露一点)正侧面。女人跪地,上身直。男人边缘只露 黑色四角裤和一小截腰,无脸。两人靠近,侧轮廓。女人占满画面。 禁止:男人正脸,正面大特写,未成年。
11. 例子|一致性短句(每条末尾粘)与参考图为 同一人,五官发型不变,不换脸。同一场景、同一衣服、同一机位,只改动作。画风一致。
12.用法:先贴 11 做锁,再套 9 或 10。Grok 出 480,剪映拉 2K。
这下面也有不用先提示词的工具喜欢的兄弟可以看看
一键生成无审查短剧:
reurl.cc/NOlKjk
一张图片即可生成劲爆的ai视频工具:
reurl.cc/rkDrGb
reurl.cc/yWv3LM
🤖 Made with AI
造梦CreateDream retweeted
想减少 AI 感,其实不一定要堆很多 Prompt。
我觉得最关键的是这 3 个:
① 先规定它是一张“真实手机随手拍”
不是棚拍、不是广告大片,就是普通人在地铁里用 iPhone 拍下来的照片。
② 别把皮肤修得太干净
毛孔、细小汗毛、轻微泛红、鼻梁和额头的真实反光都要保留。
皮肤越“完美”,反而越像 AI。
③ 保留真实的不完美
头顶飞发、凌乱发丝、轻微左右不对称、自然阴影、随意构图,都不要修掉。
这次我特意把 毛孔 + 皮肤反光 + 发丝 + 手机广角感 都加强了。
要追求“真的像随手拍了一张”。
🤖 Made with AI
一张地铁里的 AI 自拍,
随手 iPhone 自拍的感觉,
越真实,流量越高👀
Prompt🔽
一位成年东亚女性,在日本通勤电车车厢内拍摄超近距离手机自拍。人物拥有浅金色长直发、轻薄空气刘海,左侧佩戴深蓝色花朵发夹和银色爱心耳饰,穿浅紫色细肩带吊带上衣。
镜头从略低于面部的位置近距离向上拍摄,手机广角自拍视角,脸部占据画面大部分区域,头部轻微低下,下巴自然收住,眼睛平静直视镜头,嘴唇自然放松,冷淡自然表情。
严格保留真实皮肤质感:脸颊、鼻翼、额头清晰可见自然毛孔和细微皮肤纹理,轻微肤色不均,鼻梁、鼻尖、额头和嘴唇存在真实油脂反光与环境光高光,不磨皮,不美颜,不塑料皮肤。眼睛、睫毛、嘴唇纹理真实自然。
头发具有大量清晰单根发丝,头顶有自然飞发,刘海之间存在真实缝隙和细碎发丝,发丝受到车厢顶灯照射产生自然高光。
背景为真实现代日本电车内部,灰白色金属车顶、空调出风口、长条荧光灯、车窗和少量车厢结构清晰可见。冷白色车厢顶灯从上方直接照射人物,在额头、鼻梁和脸颊形成明显但自然的高光与阴影。
整体效果像 iPhone 原相机随手自拍,真实手机摄影,轻微广角透视,真实曝光,高清细节,生活感强,未经精修,没有摄影棚感,没有商业人像感,没有 AI CG 感。
photorealistic, ultra realistic skin texture, visible pores, individual hair strands, natural facial details, realistic smartphone selfie, soft indoor lighting, warm ambient light, shallow depth of field, high dynamic range, natural color grading, subtle sensor texture, highly detailed, 4K portrait photography
造梦CreateDream retweeted
We’re launching DeepSeek-V4-Pro today! 🚀
🔷 Major Agent upgrades with strong production gains!
🔷 Flexible reasoning effort for V4-Pro & V4-Flash: low for simple tasks, high for daily Agent workflows, max for complex tasks.
🔷 Native OpenAI Responses API support, optimized for Codex with one-click setup.
V4 Pro is now available on app/web. Try it via “Expert Mode”.
V4 Pro is also available via API. Model names remain unchanged—please refer to the API docs for setup details.
造梦CreateDream retweeted
DeepSeek V4 Pro 0813 is live on OpenRouter.
@deepseek_ai reports large agent gains over V4 Pro Preview: DeepSWE 62.7 (+49.9), CyberGym 83.3 (+30.6), NL2Repo 61.5 (+23.0), and Terminal Bench 2.1 87.9 (+15.8)
More providers coming online soon
Use it now: openrouter.ai/deepseek/deeps…
造梦CreateDream retweeted
我整理了一些,GROK提示词,和AI图片生成,一些我常用的姿势人物图片提示词,如果你觉得有用记得给我点个关注,或者赞,以免以后你找不到我
一、基础性感姿势(通用高张力)1. 低角度仰拍·跪姿挑衅
姿势:双膝跪地,膝盖稍分开,上身微后仰,一只手撑地,另一只手轻抚胸口或锁骨,眼神直视镜头,微微张嘴。
关键镜头:极低角度(贴近地面)从下往上仰拍,夸张透视强调胸、腰、胯。
提示词骨架:
extreme low angle shot from the ground looking up, beautiful woman kneeling on floor, knees slightly apart, arched back, one hand on floor, other hand touching her chest, intense eye contact with camera, slightly open mouth, heavy breathing expression, dramatic perspective emphasizing cleavage and hips, sensual lighting, highly detailed skin texture
2. 过肩回眸·翘臀特写
姿势:侧身背对镜头,大幅回眸看镜头,一只手撑墙或扶大腿,臀部高高翘起并向镜头扭动。
关键镜头:中近景,镜头略低于臀部,侧后方45度拍摄,焦点在翘臀与回眸眼神。
提示词骨架:
over-the-shoulder shot, beautiful woman looking back at camera over her shoulder, arched back, hips pushed toward the camera, seductive eye contact, slight smile, low angle focusing on rounded hips and waist, soft rim lighting, highly detailed
3. 俯拍分腿坐姿
姿势:坐在床边或椅子,双腿自然分开,上身微前倾,双手撑大腿内侧,低头向上看镜头。
关键镜头:高角度俯拍(45-60度),从正上方偏前往下看,强调打开的腿部与胸部。
提示词骨架:
high angle looking down, beautiful woman sitting with legs spread, leaning forward, hands on inner thighs, looking up at camera with seductive eyes, soft lighting from above, detailed body curves, intimate atmosphere
4. 贴镜近距离·喘息特写
姿势:上半身几乎贴镜头,头部微侧,嘴唇几乎碰到镜头,眼神半闭,一只手扶脖子或锁骨。
关键镜头:超近距离特写(脸+胸口),轻微畸变,营造呼吸扑面感。
提示词骨架:
extreme close-up, face and upper chest almost touching the camera, beautiful woman with half-closed eyes, slightly parted lips, soft breath, one hand on her neck, intense intimate lighting, skin texture highly detailed, shallow depth of field
5. 侧躺慵懒·长腿延伸
姿势:侧躺呈S型曲线,一条腿弯曲抬起,另一条腿伸展,一只手撑头,另一只手放腰侧。
关键镜头:低角度侧拍,镜头贴近床面,从脚部方向往上延伸,长腿主导画面。
提示词骨架:
low angle side shot along the bed, beautiful woman lying on her side in S-curve pose, one leg bent, other leg extended toward camera, hand propping her head, soft sensual lighting, long legs dominating the composition, elegant and erotic
6. 站立靠墙·双手举高
姿势:背靠墙,双手举过头顶交叉按住墙面,身体微前倾,胯部向前顶,眼神下垂看镜头。
关键镜头:低角度仰拍+轻微广角,从下往上拍全身,突出拉长身体线条。
提示词骨架:
low angle full body shot, beautiful woman leaning against wall, arms raised above head crossed, hips pushed forward, looking down at camera, dramatic upward perspective, sensual tension, detailed anatomy
二、性爱模拟姿势(单人展示)7. 模拟后入式·跪趴翘臀
姿势:双膝跪地,双手或肘部着地,腰部下塌、臀部高高翘起并向后送,可回眸看镜头。
关键镜头:极低角度正后方或侧后方,镜头几乎贴地,焦点锁在翘臀与腿根。
提示词骨架:
extreme low angle from behind, beautiful woman on all fours, knees apart, back arched deeply, hips pushed high and back toward the camera, looking over her shoulder with seductive half-closed eyes, slightly open mouth, heavy breathing, dramatic perspective focusing on rounded hips and thighs, sensual lighting, highly detailed skin
8. 模拟骑乘位·跨坐下沉
姿势:双腿分开跪坐或半蹲,身体微前倾,一只手撑身前,另一只手抚胸或扶腰,胯部下沉。
关键镜头:低角度仰拍,从正前方偏下往上拍,强调打开双腿与下沉胯部。
提示词骨架:
low angle looking up, beautiful woman in cowgirl position, legs spread wide, kneeling or squatting, body leaning slightly forward, one hand on the ground, hips lowered, intense eye contact with camera, sensual expression, emphasizing thighs and waist, dramatic upward perspective
9. 模拟传教士·仰躺抬腿
姿势:仰躺,双腿抬起并分开(可一手抱一条腿),另一只手放胸口或身侧,腰部微抬。
关键镜头:高角度俯拍,从正上方偏腿部往下看,双腿占据主要画面。
提示词骨架:
high angle looking down, beautiful woman lying on her back, legs raised and spread, one hand holding her thigh, arched lower back, looking up at camera with seductive eyes, soft lighting from above, intimate composition focusing on open legs and body curves
10. 模拟侧入·侧躺抬腿
姿势:侧躺,下侧腿伸直,上侧腿高高抬起并用手轻扶,身体呈S型。
关键镜头:低角度侧拍,镜头贴近床面,从腿部延伸方向拍摄。
提示词骨架:
low angle side shot, beautiful woman lying on her side, bottom leg extended, top leg raised high and held by hand, S-curve body, looking toward camera with half-closed eyes, soft sensual lighting, long legs and hips dominating the frame, erotic atmosphere
11. 模拟站立抱起/抵墙·靠墙抬腿
姿势:单脚站立或踮脚,另一条腿高高抬起并用手抱住膝盖/大腿,背部靠墙,身体微前倾。
关键镜头:低角度仰拍全身,略带广角,从下往上拍抬起的腿与拉伸线条。
提示词骨架:
low angle full body shot, beautiful woman leaning against wall, one leg raised high and held by hand, body arched, looking at camera with intense expression, dramatic upward perspective emphasizing legs and torso, sensual tension
12. 模拟面对面坐姿·跨坐前倾
姿势:双腿分开坐在边缘,身体大幅前倾,双手扶大腿或向后撑,胸口贴近镜头高度。
关键镜头:中近景平视或微低角度,极近距离,营造压迫感。
提示词骨架:
medium close-up low angle, beautiful woman sitting with legs spread wide, leaning far forward, hands on thighs, intense eye contact, slightly open mouth, intimate distance, focusing on cleavage and open posture, soft dramatic lighting
三、进阶高刺激姿势13. 模拟深后入·额头贴地跪趴
姿势:双膝跪地,前臂或双手贴地,额头几乎贴地,腰部极力下塌,臀部高高抬起并向镜头后送。
关键镜头:极低角度正后方近距离,臀部占据绝大部分画面。
提示词骨架:
extreme low angle from directly behind, close-up, beautiful woman on all fours, forehead almost touching the ground, back deeply arched, hips pushed high and back toward the camera, intense focus on rounded buttocks and thighs, dramatic perspective, sensual lighting, highly detailed skin texture
14. 模拟女上位前倾·双手撑“对方”
姿势:双腿大幅分开跪坐,身体大幅前倾,双手向前撑,头部低垂看镜头,胯部下沉。
关键镜头:低角度仰拍,从正前方偏下往上拍,双腿形成框架。
提示词骨架:
low angle looking up, beautiful woman in cowgirl pose, legs wide apart, leaning far forward, hands reaching forward as if supporting herself, looking down at camera with seductive eyes, hips lowered, emphasizing cleavage and open thighs, dramatic upward view
15. 模拟侧入抬腿·高难度侧躺
姿势:完全侧躺,下侧腿伸直,上侧腿被手高高拉起贴近胸口,身体扭转,臀部朝镜头微送。
关键镜头:低角度侧后方,贴近床面,从抬起腿与臀部方向延伸。
提示词骨架:
low angle side-rear shot, beautiful woman lying on her side, bottom leg extended, top leg raised high and pulled toward her chest by hand, body twisted, hips angled toward camera, half-closed eyes, soft lighting, long legs and hips dominating the composition
16. 模拟站立后入·双手扶墙翘臀
姿势:面向墙壁站立,双手高高撑墙,上身前倾,腰部下塌,臀部用力向后顶,一条腿可微抬。
关键镜头:低角度侧后方或正后方仰拍,从下往上拍完整腰臀曲线。
提示词骨架:
low angle from behind and slightly to the side, beautiful woman standing leaning against wall, hands high on the wall, back arched, hips pushed strongly backward toward the camera, one leg slightly lifted, looking over shoulder, dramatic upward perspective, sensual tension
17. 模拟传教士抬双腿·抱腿特写
姿势:仰躺,双手分别抱住大腿根部,双腿向两边大幅分开并向上抬,腰部微抬。
关键镜头:高角度俯拍,从正上方偏腿部往下看,双腿完全打开。
提示词骨架:
high angle looking down, beautiful woman lying on her back, both legs raised and spread wide, hands holding her thighs, lower back slightly arched, looking up at camera with wet eyes, intimate overhead composition focusing on open legs and body
18. 模拟面对面坐姿深插·胸口贴近
姿势:坐在边缘,双腿尽可能分开,身体大幅前倾,胸口几乎贴近镜头高度,双手扶膝盖或向后撑。
关键镜头:中近景微低角度,极近距离,面对面压迫感。
提示词骨架:
medium close-up slight low angle, beautiful woman sitting with legs spread extremely wide, leaning far forward, chest almost filling the frame, intense eye contact, slightly open mouth as if breathing heavily, intimate and dominant composition
#giok提示词 #ai提示词 #成人提示词 #GROK成人提示词
#AI动作提示词 #ai视频生成
造梦CreateDream retweeted
SpaceXAI's Grok 4.6 scores 61 on the Artificial Analysis Intelligence Index, joining the frontier in line with GPT-5.6 Sol, with standout agentic performance at lower cost
Grok 4.6 gains 5 points over Grok 4.5 on the Intelligence Index just over one month after its release, or +23 points compared to Grok 4.3. This brings SpaceXAI back to the intelligence frontier alongside OpenAI, behind only Anthropic.
Key takeaways:
➤ Grok 4.6 joins the frontier of the Artificial Analysis Intelligence Index: It scores 61, in line with GPT-5.6 Sol (max), behind Claude Opus 5 (max, 63) and Claude Fable 5 (max with fallback, 62), and just ahead of Kimi K3
➤ Strong agentic performance: Grok 4.6 achieves a GDPval-AA v2 Elo of 1753, behind only Claude Opus 5 and with overlapping confidence intervals with Claude Fable 5 and Qwen3.8 Max. It scores 50.7% on 𝜏³-Banking, among the top two scores alongside Qwen3.8 Max (51.3%), and 88.4% on Terminal-Bench v2.1, in line with the leading models
➤ Frontier-level intelligence at lower cost: Headline pricing is unchanged from Grok 4.5 at $2/$6 per 1M input/output tokens, 60%+ below Claude Opus 5 ($5/$25) and GPT-5.6 Sol ($5/$30). It cost $0.84 per task, the same as Kimi K3 with slightly higher intelligence, placing it on the Intelligence vs. Cost per Task Pareto frontier
➤ Grok 4.6 sits at Fable 5-tier on AA-Briefcase, our private benchmark of long-horizon agentic knowledge work tasks, with an Elo of 1577 - behind the Claude Opus 5 family. It is notably turn-efficient, completing tasks in ~53 turns and ~0.5B input tokens on average vs. ~103 turns and ~2.0B input tokens for Claude Opus 5 (max)
Other model details:
➤ Context window of 500k tokens (unchanged from Grok 4.5)
➤ Pricing of $2/$6 per 1M tokens of input/output; cache hits discounted to $0.5 per 1M tokens, an increase over Grok 4.5’s $0.3 per 1M tokens for cache hits
Congratulations to @spacexai and @elonmusk on the release!







































