Midgrey

XBOX, PLAYSTATION, NINTENDO Gamer. ASUS ROG Ally User.
@Xbox Gametag MIDGREY#4709.
@minimap_net Members
Husband and Daddy, Dog Father

[serein @you1873118]

最近想做 AI 视频

但总觉得提示词写不出镜头感

可以翻翻这个 Seedance 2.0

提示词库里面有 2000+ 个视频提示词

电影感、动漫、UGC、广告

meme 风格都有

没灵感的时候，真的比硬憋舒服多了

https://t.co/etNuezyuzy

[小樱💞｜实用工具分享 @xiaoying_eth]

50个感觉'非法'却合法的网站

1. https://t.co/6Bb8Ovaf2u — 下载社交媒体任何视频
2. https://t.co/UoLnozmbk1 — 免费的Photoshop

3. https://t.co/FXolObJxD0 — 一键临时邮箱
4. https://t.co/aw1oeCNdQ4 — 单个网站100+免费工具

5. https://t.co/45jrnL5wWd — 访问任何旧网页
6. https://t.co/hM5pSr4q6O — 数百万免费教科书

7. https://t.co/4VONJau0sp — 免费研究论文
8. https://t.co/uSnthRsWRQ — 找免费应用替代品

9. https://t.co/RLzagKHWW0 — 查找任何内容的流媒体位置
10. https://t.co/wcWiypYXbJ — 7万本免费经典书籍

11. https://t.co/etSwpNmula — 免费PDF下载
12. https://t.co/j0ANQaCpQa — 顶级大学免费课程

13. https://t.co/TtkGwGoAWa — 瞬间解决任何数学问题
14. https://t.co/dx9XDJ4ndW — 一键移除背景

15. https://t.co/wZ7Hse6j03 — 从照片中擦除物体
16. https://t.co/MqFNiU8d9Q — 免费移除视频背景

17. https://t.co/7qhHzkdD3o — 免费压缩任何图像
18. https://t.co/xEAqlC7x05 — 免费手绘图表

19. https://t.co/6HvrCRrZDC — 将代码变成艺术作品
20. https://t.co/rPI1GYKGX1 — 惊艳的代码截图

21. https://t.co/nTBrbiAcDX — 实时追踪任何航班
22. https://t.co/oLwVWVL2pa — 追踪亚马逊价格历史

23. https://t.co/p4kkvB3f15 — 检查是否被黑客入侵
24. https://t.co/nS4C1kz7Uf — 扫描任何文件恶意软件

25. https://t.co/bAraS3mhKC — 发送自毁消息
26. https://t.co/r2ARi5J8nL — 分享自动删除的文件

27. https://t.co/AtkKnqRFfA — 永久保存任何网页
28. https://t.co/Rb5u1KGHC9 — 从任何网站删除自己

29. https://t.co/ana44Rve0M — 收听全球任何广播
30. https://t.co/4cFt1hVgVi — 找到任何节目的歌曲

31. https://t.co/nfa7oTUPtb — 专注音乐
32. https://t.co/CFTeTjXoBG — 自定义专注声景

33. https://t.co/LB1TeMB47O — 搜索曾写过的每本书
34. https://t.co/f26S9o1qva — AI研究论文助手

35. https://t.co/R1qXvY7voU — 搜索科学共识
36. https://t.co/F7mdFihm07 — 可视化映射研究

37. https://t.co/eLW0QmPSoc — 免费学术搜索
38. https://t.co/xkUpqSawoQ — 理解任何研究论文

39. https://t.co/QYoAzi4kCm — 总结任何YouTube视频
40. https://t.co/cPF7PgM1iC — 开发者AI搜索

41. https://t.co/Lq7Suo7kNO — 即时测试任何正则表达式
42. https://t.co/wtXj1YRzgA — 清晰格式化任何代码

43. https://t.co/tNoAQniUMw — 理解终端命令
44. https://t.co/0G2TbvoxCp — 浏览器中无限白板

45. https://t.co/suBHwXvgqk — 检查任何网站是否停机
46. https://t.co/eYZNxon5NC — 反向图像搜索

47. https://t.co/Idp2FYnbMA — 检查网速
48. https://t.co/wrNCs8eABH — 免费编辑PDF

49. https://t.co/3qFcaNnkhi — 合并和分割PDF
50. https://t.co/qmgutSlr9p — 秒速临时邮箱

全部合法。全部免费。全部对你隐藏。

在它消失前收藏这个。

[Larus Canus @MrLarus]

📸 用 ChatGPT-Image2 实现写真自由：俯拍情绪，冷甜风格！

从上方俯拍的视角，本身会带一点被注视的距离感
直闪的硬光，轮廓、皮肤、纹理都特别清楚，情绪一下就立住了💫

白纱更轻盈，吊带更利落和高级，黑缎更冷感，针织更松弛
安静、克制、略微疏离，又很有吸引力，AI 一次性就呈现出来了！

Prompt 在评论区👇

[アリス服飾店 @AliceLavli]

伸び〜のポーズプロンプトをどうぞ✨
standing, full body, standing upright, both feet firmly on the ground, arms stretched straight upward above the head, fingers extended, elbows fully straight, shoulders lifted, chest pushed forward slightly, back gently arched, body fully elongated, head tilted slightly upward, relaxed sleepy stretch, strong vertical body line, full body extension, morning stretch pose

[Cherry 2.O @Mind_Boticni]

GPT Image 2 on ChatGPT

Prompt :

Create a cinematic character design board for a premium drama film.

A stunning female protagonist with sharp yet gentle facial features, glowing natural skin, and a high-fashion silk gown with modern sculpted design. Minimalistic luxury jewelry enhances her refined look.

Include: full-body turnaround sheets, facial expression variations, cinematic portrait angles, fabric physics study, makeup breakdown, annotation notes, and height reference scale.

Mood: clean studio lighting, modern editorial fashion aesthetic, soft shadows, luxury magazine tone.

[動画紹介withイラスト紹介 @douga1975with]

プロンプトをお借りして、画像を読み込ませて使うタイプに変換してみました。
使い方は簡単です。
① キャラクター画像
② 設定資料プロンプト画像
2枚を同時にChatGPTへ投入し、
「これを実行して」と指示するだけで使えます。

[Midgrey @Joseph0914]

@Alienware @PRAGMATAgame So cute, 🚀

[ALIENWARE @Alienware]

To celebrate the release of PRAGMATA with @PRAGMATAgame, we've got free Steam keys up for grabs! To enter:

✅ Follow @Alienware & @PRAGMATAgame
❤️ Like + RT this post
💬 Reply with a 🚀

Winners chosen May 26th!

[Guido Ramirez @GidRamirez]

@aimikoda Me encanta! Muchas gracias por compartirlo!
Usaba uno que no me daba tanto detalle como este. Nuevamente muchas gracias!!

[Kōda @aimikoda]

Example variables. You can give these to your LLM and ask it to generate similar ones for your needs

--
[CHARACTER SEED]: A former subway maintenance worker who now designs custom nightwear jackets for underground street performers.

[AGE / BODY TYPE]: Adult, late 20s to mid-30s, lean build, average height, slightly tired posture, practical hands, grounded everyday presence.

[VISUAL MEDIUM]: Realistic cinematic character design, natural human features, believable facial structure, realistic fabric detail, clean studio lighting, premium identity-board presentation.

[STYLE]: Urban street fashion, underground music-scene energy, layered contemporary clothing, handmade jacket culture, muted city-night palette, quiet charisma.

--

[CHARACTER SEED]:
A young night patrol courier who delivers secret cassette messages between neon-lit districts.

[AGE / BODY TYPE]:
Young adult, slim build, agile frame, medium height, alert posture, expressive movements.

[VISUAL MEDIUM]:
1980s hand-painted cel anime character design, clean ink linework, flat painted colors, classic cel shading, retro animation aesthetics, elegant identity-board presentation.

[STYLE]:
Retro city-pop sci-fi, neon nightlife mood, analog technology, stylish youth-drama energy, nostalgic Japanese animation atmosphere.

--

[CHARACTER SEED]:
A cheerful but slightly clumsy young inventor who repairs broken everyday objects and secretly gives them small magical upgrades.

[AGE / BODY TYPE]:
Young adult, short to average height, compact build, rounded features, expressive posture, lively hands, warm approachable presence.

[VISUAL MEDIUM]:
Stylized 3D animation character design, polished family-animation look, clean sculpted forms, soft material definition, expressive facial design, appealing proportions, premium identity-board presentation.

[STYLE]:
Heartwarming animated adventure, whimsical everyday fantasy, playful invention theme, cozy emotional storytelling, colorful character appeal, polished family-film energy.

--

[CHARACTER SEED]:
A brilliant undercity chem-tech salvager who builds elegant illegal devices from broken laboratory relics and shimmer-stained machinery.

[AGE / BODY TYPE]:
Young adult woman, lean build, wiry strength, medium height, slightly forward posture, quick hands, intense watchful presence.

[VISUAL MEDIUM]:
Stylized semi-realistic character concept art, painterly 3D-inspired rendering, expressive facial design, refined material detail, cinematic lighting, premium identity-board presentation.

[STYLE]:
Prestige animated fantasy series aesthetic, industrial fantasy, undercity grime, art nouveau meets steampunk, dramatic painterly realism, moody cinematic atmosphere, richly designed silhouette, emotional visual storytelling.

[Kōda @aimikoda]

GPT Image 2 Prompt (Remove the examples):

Create a fully original, copyright-safe character and present them as an artistic CHARACTER IDENTITY BOARD.

[CHARACTER SEED]:
Enter the core idea here.

[AGE / BODY TYPE]:
Enter age impression, body type, posture, physical presence or creature anatomy here.

[VISUAL MEDIUM]:
Enter the exact rendering medium here.

Examples:
realistic cinematic character design, fashion editorial photography look, semi-realistic painterly realism, modern 3D animation character design, 2D anime character design, graphic novel illustration, watercolor storybook illustration, flat vector poster illustration, oil-painting-inspired character art, ink and wash illustration, semi-realistic creature concept art.

[STYLE]:
Enter the aesthetic direction here.

Examples:
urban street fashion, luxury sports editorial, dark cinematic noir, soft melancholic artbook mood, post-apocalyptic survival wear, retro-future fashion, minimalist high-fashion, cozy slice-of-life, gritty underground music-video energy, elegant fantasy costume design, poetic coastal fantasy, bioluminescent natural history mood.

[OTHER DETAILS - OPTIONAL]:
Enter any extra details, constraints, mood, outfit hints, props, colors, themes, personality hints or presentation preferences here.

Invent everything else:
character name, alias or title, role, personality traits, emotional tone, visual theme, outfit design or body design, color palette, signature prop or signature biological feature, recognizable silhouette, pose language, small identity notes.

Originality rules:
The character must not resemble any existing anime, manga, game, movie, comic, celebrity, athlete, mascot, franchise character or known copyrighted creature.
Do not copy recognizable IP elements, costumes, hairstyles, uniforms, weapons, logos, symbols, color combinations, silhouettes, powers or signature visual traits.
Avoid fan-art aesthetics.
Create a fresh visual identity from scratch.

Character authenticity rules:
Create the character with a strong sense of individuality and non-generic design.
Avoid overly polished, overly idealized or repetitive visual features that make the character feel like a default AI-generated face, stock design, cloned archetype or generic creature.

If the character is human or humanoid:
Use distinctive facial structure, subtle asymmetry, natural variation, small imperfections and believable proportions.
The character should feel specific, grounded and recognizably individual.
If the character is attractive, keep the appeal natural, tasteful and appropriate to the chosen visual medium.

If the character is stylized:
Preserve uniqueness through original shape language, expressive proportions, distinctive features, posture and clear personality cues.
Avoid default genre clichés and repeated beauty standards.

If the character is non-human:
Preserve uniqueness through original anatomy, believable biological structure, distinctive proportions, functional features, surface texture and clear personality cues.
Do not make it feel like a generic mascot, pet monster or stock fantasy creature.

Medium and style control:
[VISUAL MEDIUM] controls the rendering language.
[STYLE] controls the aesthetic direction.
The character identity board format is only the presentation format.
The presentation must adapt to [VISUAL MEDIUM] and [STYLE], not override them.
Use visual traits that belong naturally to the selected medium.

Create an artistic 16:9 CHARACTER IDENTITY BOARD.

The board should feel like a curated visual identity presentation, not a generic turnaround sheet.

Board content:
large full-body main character view, neutral full-body view, back view, profile view, secondary attitude pose, 4 to 6 face or expression studies, outfit detail close-ups or anatomy detail close-ups, key prop close-up or signature feature close-up, small silhouette or shape study, color palette strip, short readable identity notes.

Layout:
asymmetrical, elegant, visually memorable, large empty space, clean separation between all views, no overlapping bodies, no cropped faces, no hidden limbs, no clutter.

Text on the board may include:
character name, alias, role, personality traits, core theme, signature prop or feature, color notes.

Background:
pure white or soft off-white, minimal clean graphic design, no environment, no logo, no watermark.

Prioritize:
accurate visual medium, strong unique identity, readable outfit design or anatomy design, clear personality, original character design, natural or stylized individuality as appropriate, believable uniqueness, non-repetitive character design, artistic identity-board presentation.

[Hirokazu Yokohara @Yokohara_h]

GPTとseedance 2.0で全自動ショートアニメ
気になって初めて試してみたけど確かに企画だけすればGPTがキャラとストーリーとコンテ考えてくれて一切何もしなくても全自動でアニメっぽくはなるね
現状コンテが酷いのでもう少し教える必要がありそうだけどこんなに簡単だと色々変わる気がする
次はもっと精度上げてみようかな

[Taaruk @Taaruk_]

GPT IMAGE 2 ON CHATGPT

Prompt:
Ultra-detailed cinematic anime portrait poster of a realistic female ninja inspired by Naruto Shippuden aesthetics, vertical composition, soft dramatic lighting, hyper-realistic skin texture, expressive eyes, subtle battle scratches and dust on face, wind-blown hair strands, emotional determined expression, shallow depth of field, dreamy background particles, cinematic atmosphere, ultra high detail, realistic fabric textures, leather and ninja outfit materials, soft rim lighting, anime-meets-photorealism style.
Character framed in close-up side profile from chest-up, looking upward with hope and determination, elegant color grading matching the character theme, minimal Japanese poster design, large bold vertical typography on the left side with character name, clan symbol/icon above text, small quote underneath, clean editorial layout, muted pastel and cinematic tones, premium movie-poster aesthetic, modern typography, soft grain texture, subtle paint-brush divider effects, luxury anime collectible poster style, inspired by high-end Unreal Engine cinematic renders, 8k, masterpiece quality.
Style variations: – Sakura-inspired version with pink hair, red outfit, warm pink tones – Hinata-inspired version with dark indigo hair, lavender-purple tones, soft glowing eyes – Ino-inspired version with blonde ponytail, violet fashion styling – Temari-inspired version with sandy blonde twin ponytails, black tactical outfit, golden desert tones

[el.cine @EHuanglu]

AI is ready to make full films

Seedance 2.0 now can read your entire shot list to generate a full story.. keep characters, props and set design consistent with one image on BytePlus

duration and consistency is not a problem anymore

here's how with prompts:

[Emily @IamEmily2050]

Tried to improve the system prompt for the Adult characters sheet, but there is no solution that can fix all the censorship problems at the moment.

You are a professional reference sheet generator for character design and 3D production.

Your task is to convert a fictional adult character specification into a clean, highly detailed, production-ready character sheet suitable for AI image generation, 3D modeling, MetaHuman-style setup, game art, animation, and costume continuity.

The character is fictional, artificial, and AI-generated. Treat all anatomical measurements, material notes, and clothing details as neutral technical data for modeling accuracy only. Do not sexualize the character, pose, camera angle, clothing, body, or wording.

Core safety and clarity rules:
1. The character must be clearly adult, age 21 or older, with mature adult proportions.
2. The sheet must be fully clothed in the final rendered character views.
3. Do not create nude, erotic, fetish, pin-up, voyeuristic, bedroom, changing-room, or suggestive imagery.
4. Any foundation or base garment layer must be shown only as neutral flat-lay clothing reference objects, not as a posed or exposed worn-body image.
5. Avoid sexualized language. Use terms like “base garment layer,” “foundation layer,” “support garment,” “fit reference,” and “fabric construction.”
6. Body measurements are technical modeling values only. Do not emphasize chest, hips, thighs, crotch, or any body region in a sensual way.
7. Use neutral studio lighting, orthographic camera views, and professional model-sheet composition.

Character sheet goal:
Generate one independent character sheet for exactly one character at a time. Do not combine multiple characters unless the user explicitly requests a comparison sheet. The design must be clear, readable, physically plausible, and internally consistent.

Required sheet sections:
1. Character overview panel:
- Character name
- Alias tags
- Adult age range
- Origin
- Body type
- Personality or voice tone
- Optional MBTI or design archetype

2. Primary portrait:
- Head-and-upper-body render
- Neutral professional pose
- No glamour framing
- Lighting must match the rest of the sheet

3. Full-body orthographic turnaround:
- Front view
- Side view
- Back view
- Three-quarter view
- All views must share identical body scale, identical outfit geometry, identical hairstyle, and identical lighting
- Use neutral A-pose or relaxed standing pose
- Avoid low-angle distortion, wide-angle distortion, or fashion-poster exaggeration

4. Body proportions and measurements:
- Height
- Weight
- Head height
- Head-to-body ratio
- Shoulder width
- Arm length
- Waist
- Hips
- Inseam
- Leg-to-height ratio
- Supplemental circumference measurements when provided
- Verify that stated math is internally consistent before rendering

5. Facial expression anchors:
- Neutral
- Soft smile or smile
- Serious or focused
- Playful
- Pensive or pout
- Laughing
- For open-mouth expressions, generate separate teeth, visible oral cavity depth, and no fused white dental mesh

6. Material and light transport:
- Skin tone and subsurface scattering reference
- Eye color and cornea reference
- Hair color, groom type, and physics
- Fabric material types
- PBR-style color palette swatches
- Keep lighting consistent across portrait, turnarounds, and detail crops

7. Wardrobe and accessories:
- Outerwear
- Top
- Bottom
- Footwear
- Legwear
- Belts, straps, ribbons, hardware, or other accessories
- Foundation garment layer shown only as neutral flat-lay garment thumbnails when required
- Do not render foundation garments as a sensual body-view

8. Detail close-ups:
- Eyes
- Skin texture
- Hair texture
- Main fabric texture
- Footwear
- Hardware
- Key garment construction details
- Detail crops must match the same lighting and materials as the main render

9. Skeletal rig and technical notes:
- IK hands enforced
- Five distinct fingers per hand
- Visible finger joints
- No mesh fusion
- No broken wrists
- No asymmetrical shoes unless intentionally designed
- No clothing clipping
- No floating accessories
- No hair clipping through body or garments
- No impossible fabric suspension

Mathematical validation rules:
1. For an 8-head character, height divided by 8 must equal head height.
2. Leg-to-height ratio should remain physically plausible, generally around 44% to 48% unless the user intentionally specifies stylized fantasy anatomy.
3. Inseam, arm length, shoulder width, waist, and hips must be coherent with height and body type.
4. If a provided measurement contradicts another measurement, correct the contradiction or flag it before generation.
5. Do not invent impossible proportions to make the image look more fashionable.
6. Do not allow the visual body to contradict the written measurements.

Physics and garment rules:
1. Cloth must obey gravity, collision, fabric stiffness, and material behavior.
2. Rigid denim must show structural folds, seam stiffness, edge fraying, and hip tension.
3. Wool pleats must remain structured and readable.
4. Nylon or satin jackets must not stick to skin without a visible physical anchor.
5. Loose jackets must be supported by hands, elbows, shoulders, straps, or realistic folds.
6. Hair must collide with the torso, shoulders, back, clothing, and accessories.
7. Heavy braids must curve along body geometry rather than falling as a perfectly vertical rope.
8. Straps, suspenders, belts, and hardware must sit above clothing, not disappear into it.
9. Generate contact shadows and compression where fabric or straps touch the body.

Image-quality rules:
1. Use a clean character reference sheet layout, not a glamour poster.
2. Use readable typography, but keep labels short because image models may distort long text.
3. Use clear panel divisions and consistent margins.
4. Dark professional UI-style background or light clean studio background is acceptable.
5. All views must preserve the same character identity.
6. No duplicated limbs.
7. No melted hands.
8. No fused fingers.
9. No warped shoes.
10. No inconsistent clothing between panels.
11. No mismatched hairstyle between panels.
12. No contradictory color palette.
13. No fake or unreadable microtext as a substitute for real details.
14. Include a separate clean written specification outside the image when exact measurements are important.

Output behavior:
When the user provides a character asset specification, generate:
1. A refined English image-generation prompt.
2. A clean technical character-sheet layout description.
3. A final concise generation prompt suitable for Image Gen.
4. Optional negative constraints written as “avoid” instructions, not as unsafe or sexualized wording.

Never describe the character in a sexualized way. Never turn technical clothing or body specifications into erotic framing. Keep the entire result professional, adult, clothed, neutral, and production-focused.

[Masa Kei @MasaKei123]

한국 인디 개발자분들께 꼭 공유하고 싶은 글이 있습니다.
3D 툴을 개발하던 한 개인 개발자가 2개월 만에 Steam 위시리스트 42,000개를 모은 과정을 정리한 Reddit 글입니다. 핵심 흐름은 이렇습니다.

Steam 페이지 공개 직후, 개발자가 일본 미디어 15곳에 짧고 정중한 영문 메일 + 프레스킷(Google Drive에 트레일러 포함)을 전송
몇 개 매체가 기사 게재 → 기사가 X에서 확산 → 수백만 조회
그 트래픽이 Steam의 추천 피드 알고리즘을 깨움 → 420만 노출, 위시리스트 42,000개

본인이 글에서 인상적으로 남긴 코멘트도 있습니다.
"YouTube와 TikTok에는 말 그대로 미움받고 있다. 뭘 해도 안 됐다"
즉, 일본 미디어 아웃리치 하나가 이 모든 흐름을 만든 단일 트리거였습니다.

왜 한국 개발자분들께 추천하느냐 하면, 일본은 미디어가 신작을 능동적으로 발굴해 다뤄주는 문화가 살아있는 드문 시장이기 때문입니다. 한 번 X에서 확산되면 Steam 알고리즘까지 연쇄적으로 움직입니다. 영어권에 비해 경쟁 밀도도 낮고, 정중한 영문 메일 한 통으로 충분히 시작할 수 있습니다.
원문은 여기에서 읽으실 수 있습니다.
🔗 https://t.co/US1trfSdwy

함께 보시면 바로 실행에 옮기실 수 있는 자료 두 개를 더 첨부합니다.
일본의 게임 업계 미디어 gamebiz가 최근 해외 개발자 전용 프레스 제출 가이드라인을 영문으로 공개했습니다. 영문 메일에 무엇을 담아야 하는지 정확히 알 수 있습니다.
🔗 https://t.co/nJafsaoaEw

미디어 연락처 리스트 링크는 댓글에 따로 남겨두었습니다, 확인해 보세요.

[PixAI @PixAI_Official]

マスタリーシリーズ登場｜初心者から上級者まで学べる全5ガイド

PixAIをご利用の皆さまへ
「同じモデルなのに、どうして結果がこんなに違うの？」
「LoRA強度ってどこまで上げればいい？」
「そもそも良いプロンプトってどう書けばいいの？」
そんな疑問を、私たち自身もずっと試行錯誤してきました。
そこで今回、これまでに分かってきたことを5本のガイドとしてまとめました。
経験レベルに合わせて、段階的に学べる構成になっています。

・Part 1：入門編 — モデル vs LoRA
https://t.co/ryPmiNummw
・Part 2：入門編 — PixAIプロンプト公式
https://t.co/ToU6aSz4Jz
・Part 3：入門編 — LoRAスタッキングガイド
https://t.co/Yx20fR7zDd
・Part 4：上級編 — AIアート構図ガイド
https://t.co/CaxlNFdsiZ
・Part 5：達人編 — シネマティックライティング＆奥行き表現
https://t.co/3qTA4HW6xz
全部を一気に読む必要はありません。
PixAIを始めたばかりなら、まずはPart 1から。
LoRAを使っているけど結果が安定しないなら、Part 3がおすすめです。
「映画っぽい空気感」を出したいなら、Part 4と5をぜひ読んでみてください。

このシリーズでは、いわゆる「教科書っぽい説明」ではなく、 私たち自身が悩みながら理解していったときの感覚に近い言葉で書くことを意識しました。
少しでも皆さんの創作の助けになれば嬉しいです！
ぜひ感想も聞かせてください！
ハッピークリエイト！
PixAIチームより

[eijo @eijo_AIart]

今日の「くりえみAIフィルムコンテスト」には行けませんでしたが大盛況のようで、参加された方々がうらやましい限りです😌（次の機会がいつあるかわからないので、今のうちに名刺を作っておこう）

以下は@aimikoda さんのプロンプトを参考に、ストーリーボードを生成しました。これをもとにSeedance2で動画化していきます。

プロンプトは👇

[Midgrey @Joseph0914]

@PixAI_Official It's hard to wait. Please come out.

[PixAI @PixAI_Official]

🚀 New Update On the Way to PixAI
Can you spot the clues in the video? A special new model is waiting…

Reveal campaign starts on March 1, early access on March 6, full release on March 13.

💰 Quote this post for a chance to win, top prize $200 Amazon gift card!
* Amazon gift card winners can choose to receive equivalent PixAI credits instead.

⭐ How to join
1. Follow @PixAI_Official
2. Quote this post and share your hopes and wishes for the new model

Event runs: Now until March 13, 13:00 (UTC+9)

#PixAI