Loading…
Loading…
Loading the case library.
Loading…
Case library · Public evidence
I think AI Agents are about to completely change how we prompt video models. I gave the MiniMax Design AI Agent one imag
Prompt
I think AI Agents are about to completely change how we prompt video models. I gave the MiniMax Design AI Agent one image and one simple instruction: “Act as an expert FPV director and…
Curious Fox & Mountain Stream Vlog with Seedance 2.5 Prompt: Create a single continuous self-filmed vlog recorded on a h
Prompt
Create a single continuous self-filmed vlog recorded on a handheld action camera in 16:9 aspect ratio. The camera is always held by the anthropomorphic fox, creating a natural first-person…
Some people still think AI does't look real - This video just ended the argument Made with Seedance 2.5 1080p on maxfusi
Prompt
Raw vertical phone video (UGC), handheld throughout, filmed by a friend who walks with the group, casual sway, natural night lighting, wide-angle lens, deep focus. Characters: The hero is…
Seedance 2.5 Prompt: Main subject: young Indonesian woman, early 20s, natural everyday appearance, faded mustard-yellow
Prompt
Main subject: young Indonesian woman, early 20s, natural everyday appearance, faded mustard-yellow oversized cotton t-shirt, brown high-waisted linen wrap skirt reaching just below the…
Made with seedance 2.0 Prompt: Create a cinematic, photorealistic lifestyle video of a young East Asian woman in a cozy
Prompt
Create a cinematic, photorealistic lifestyle video of a young East Asian woman in a cozy modern apartment by the sea. The video has a warm, natural morning atmosphere with soft daylight…
A comprehensive video prompt for Seedance 2.5 designed to generate a 30-second water obstacle competition in the style of a Japanese variety show, emphasizing character consistency and complex camera work.
Prompt
Generate a 30-second, 16:9 widescreen, highly realistic live-action large-scale water obstacle competition show. The overall presentation should have the live-broadcast feel of a major…
The author uses Nano Banana for commercial portrait retouching of an uploaded photo. The core workflow locks face identity and proportions and rebuilds only soft studio lighting, a pure white background and natural skin texture, with explicit negative constraints against changing the face.
Prompt
Preserve the person's face exactly as it appears in the uploaded image. Do NOT regenerate, reconstruct, reinterpret, beautify, reshape, or replace the face. Maintain the exact: Facial…
The author first makes environment and character reference images in Midjourney, then has MiniMax H3 generate a 15-second seamless-loop 360-degree turntable animation of a shipwright. The prompt controls scene assembly, text animation, transitions, music and sound beat by beat.
Prompt
REFERENCE ROLES Image 1 defines the environment architecture only — arched timber shipyard framework, hanging pulleys, ropes, beams and tool fragments. Reinterpret entirely as bold ink…
A text-to-video title design generating a 15-second opening sequence for THE JADE SUN, a fictional Mayan-themed series, with a restricted material palette, typeface and sound, and an explicit ban on mixing elements of other civilizations. The author is a creative director at Prime Video.
Prompt
Create a visually rich 15-second, 16:9 opening-title sequence for a historical drama titled THE JADE SUN, set within a powerful Classic Maya city-state. The story concerns royal…
An entrance template with a character reference image, moving in four stages from detail close-up and back-view medium shot to a face portrait and a full-body hero shot. Swap in a different character to reuse it. The original post drew 93K+ views.
Prompt
Use @[char ref] as the sole character reference. Preserve the exact identity, face, body proportions, hairstyle, outfit, colors, materials and overall silhouette of the character…
Five three-second Paris street shots anchored to named landmarks, with one pink handbag repeated in every scene so the model has an identity anchor to hold onto through rapid cuts.
Prompt
A 15-second hyper-realistic cinematic fashion campaign video. European woman, fair glowing skin, sharp features, wavy blonde hair in sleek high ponytail, small dark rectangular sunglasses,…
A web case by @felixhhaas using Lovable, including the public result, full prompt, and original source.
Prompt
Most companies have messy email signatures. So I built a tiny app for Lovable that fixes it 🔥 Everyone gets: → A consistent, on-brand signature → A live preview → One-click copy for…
A website case by @viktoroddy created with Nano Banana · Flow · AntiGravity, including the public result, full prompt, and original source.
Prompt
Build a premium, high-end hero section for a video editing agency named 'Logoisum' with the following specifications: Background: Implement a full-screen, looping video background using…
A web case by @viktoroddy created with Gemini 3.1, including the public result, full prompt, and original source.
Prompt
Build a premium electric vehicle landing page with a dark, cinematic aesthetic. The entire site uses a black background (hsl(0,0%,0%)) with light text (hsl(0,0%,95%) as primary) and orange…
A casual smartphone-style family video prompt about a sunny beach day trip, featuring jump cuts and realistic handheld phone shake.
Prompt
Super casual real smartphone home video footage of a sunny beach day outing with friends. Natural mobile phone camera recording with slight authentic handheld shake, normal frame r
A professor writes and explains a trigonometric proof on a traditional chalkboard. The case is useful for testing legible writing, instructional pacing, and continuity between speech and equations.
Prompt
A professor writes out a mathematical proof for trigonometric identities on a traditional chalkboard, explaining the step he is currently on in the equation.
A real coding/UI case that links to the open-source 3DCellForge project, turning image input into an interactive 3D workflow.
Method / Code
View method ↗
[标题] AI漫剧现代男主霸总形象提示词（附提示词） [文案] 模型：GPTImage2.5 🕸️：灵感啵啵（可一键get同款） 提示词： 一、人物图像生成提示词（画面主体） 一名 [视觉誊录·图1·一致度0.96] AI 漫剧现代男主形象提示词 ✦ 病娇矜贵 · 斯文败类 · 暗格逸钧 ✦ 总提示词 一名 27 岁…
Prompt
一名 27 岁亚洲男性,精致瓜子脸,下颌线锋利利落,冷白奶油肌,面部病态惨白,毛孔级细腻纹理。桃花眼狭长,瞳色深棕偏黑,睫毛浓密纤长,眼神病娇、疯批,带自然卧蚕阴影。高挺窄鼻梁,鼻骨感清晰,M…
[标题] AI漫剧现代男主霸总形象提示词（附提示词） [文案] 模型：GPTImage2.5 🕸️：灵感啵啵（可一键get同款） 提示词： 一、人物图像生成提示词（画面主体） 一名 [视觉誊录·图1·一致度0.96] AI 漫剧现代男主形象提示词 ✦ 病娇矜贵 · 斯文败类 · 暗格逸钧 ✦ 总提示词 一名 27 岁…
Prompt
成熟亚洲男性，清瘦型瓜子脸，骨相凌厉锋利，亮冷白奶油肌，面部病态惨白，自带逆水寒捏脸质感、BJD精致立体面部，毛孔级细腻通透肤质。狭长微窄桃花眼，睫色深棕偏黑，睫毛浓密纤长，眼神专注深邃，自带…
Cute Korean village vibes with a little chicken chase. Simple moments, beautiful memories. Created on seedance 2.0 Create a natural, cinematic, ultra-realistic …
Prompt
Cute Korean village vibes with a little chicken chase. Simple moments, beautiful memories. Created on seedance 2.0 Create a natural, cinematic, ultra-realistic video featuring a…
[标题] AI漫剧现代男主霸总形象提示词（附提示词） [文案] 模型：GPTImage2.5 🕸️：灵感啵啵（可一键get同款） 提示词： 一、人物图像生成提示词（画面主体） 一名 [视觉誊录·图1·一致度0.96] AI 漫剧现代男主形象提示词 ✦ 病娇矜贵 · 斯文败类 · 暗格逸钧 ✦ 总提示词 一名 27 岁…
Prompt
成熟亚洲男性,锋利立体骨相, 窄长脸型,下颌线紧致利落, 冷白通透肤质,轻微病态惨白感, 毛孔级细腻皮肤纹理。狭长桃花眼, 深棕黑瞳孔,眼神冷淡疏离, 暗藏疯批压迫感,斯文败类气质, 年上大佬感。高挺窄鼻梁…
A beautiful moment, a tense conversation, and a story hidden behind every look. Cinematic AI storytelling with realistic emotions and film-like visuals. Created…
Prompt
A beautiful Korean girl with long, naturally wavy black hair, soft realistic facial features, natural skin texture and subtle makeup, wearing a clean white casual jacket over a simple gray…
He survived the first impact. The second one is coming. Created On Seedance 2.5. Prompt 👇 VISUAL STYLE: Ultra-cinematic science fiction, photorealistic CGI, da…
Prompt
VISUAL STYLE: Ultra-cinematic science fiction, photorealistic CGI, dark futuristic city at night, drama…
Bro chose violence… she chose revenge. Created on Seedance 2.5 Prompt: Create an exactly 30-second, 16:9, 24fps hyper-realistic Korean live-action action-comedy…
Prompt
Create an exactly 30-second, 16:9, 24fps hyper-realistic Korean live-action action-comedy scene inside one modern café. All characters are adults. Keep the café layout, screen direction,…