Liam Wilson

[nicco @0xnicc0]

Opus 5.5 promt: https://t.co/IFg7itKHVY

[nicco @0xnicc0]

Opus 5.5 is the new After Effects

and nobody's talking about it…

Feed the PDF below to Opus 5.5, you'll thank me later

Then start with the guide below, it's all you need for motion design in Opus 5.5

[兴雨 @YxzRainy]

卧槽😱这样发展下去，视频模型天塌了啊！

同一个提示词，Gemini 4 Argon 和 Claude Opus 5.5 分别搓出了一只在宫殿里潜行的白虎。

这毛发、水雾、光影，还有电影级的画面质感，不知道的还以为是 Veo/即梦生成的！？

众所周知，Gemini 4 Argon 不是视频模型，Claude 也没有原生视频生成能力。

所以，这他🐎到底是怎么做到的？？？ 🤯

[Movez @0xMovez]

I just solved "Robotics Engineering" education with Opus 5.5 + /colorfigures skill

no need to spend $1000+ on a starter kit to learn what a resistor does, just visit my interactive site:

• 13 real boards: Arduino Uno, L298N, Raspberry Pi 5...
• build part by part: pick & place, reflow, solder, plug in
• break them on purpose: reversed caps pop, chips smoke, LEDs burn out

send the skill to your Claude:
> https://t.co/1879rauiu7

try it free, no sign-up ↓

[花叔 @AlchainHust]

推荐一个X隐藏宝藏博主 @Nenlyonis
零代码基础设计师出身，先靠AI教程在B站做到71万粉丝；今年又用AI从零自学、独立做了一款Steam游戏，免费上线两个月，10万+玩家，好评如潮。

[Brian Pridgen, MD @pridgenmd]

@GoogleDeepMind Could you provide a link to this interactive tool? Would love to try it out.

I built something similar, presumably with some of the same open source anatomy datasets.
https://t.co/SL0KAFo8os

[Google DeepMind @GoogleDeepMind]

Gemini 4 Argon combined three open-source datasets to build an interactive 3D visualization of human anatomy.

Toggle different layers in motion, isolate muscles, or zoom inside a beating heart.

[Lentils @Lentils80]

An output from the updated Carbon checkpoint 😼

Like I said in my previous post, Carbon is noticeably better than Gemini 4 Argon at 3D (tho still not perfect). It seems like Google is currently focusing on improving that domain before moving on to frontend/UI

[Bee @thtbee_]

Gemini 4 Carbon - A Mechanical flower!

This took ~1hr 43mins fully made in three.js, I love the small bee it added which wasn't in the prompt but I did say it to be creative and didn't limit it with constraints..

A lot of improvements over Argon tbh.. I'll share my observations once i have some more outputs.. I've got one more and still trying to get more.. stay tuned!

[Sac @Saccc_c]

你可能不敢相信，下面这个3D动效视频是我用一个新上线的国产模型纯代码做出来的。视觉质感完全不输Opus 5.5，价格却便宜近30倍！！

这个模型就是MiniMax的M3.1 Flash，也是此前在OpenCode和OpenRouter上匿名发布且一度爆火的Space Bunny模型。匿名期间，模型在OpenRouter上的调用连续10天霸榜，受到开发者的广泛喜爱

根据官方公布的测评结果，它在复杂开发、computer use 等任务能力上已经接近前沿模型，token成本却极低，性价比层面可谓直接斩杀了包括DeepSeek、Kimi在内的大部分模型

为了测试它的综合能力，我用其在MiniMax Code里实测制作了一个动效视频，核心是要求模型「用纯代码形式制作，最大化视觉表现，让动效、转场达到高级宣传片质感」

经过多轮迭代，M3.1 Flash产出了下面这个自我介绍宣传片，3D动态感和整体叙事节奏都很在线。整体实操下来有2点特别惊艳我：

1、Agent编排、执行能力很强。模型可以合理分配subagent并行执行，并派独立的 subagent 审查效果，整体制作效率高

2、视觉能力很在线，懂得实时截图来查看、分析视频画面，并根据画面细节不断优化、迭代

如果你也想像opus 5.5那样用AI做视频，但苦��没有claude，M3.1 Flash完全可以作为你的完美平替！

官方目前还发布了两个超级福利：10月25日前，所有MiniMax Code 订阅用户可以不限量使用 M3.1 Flash；10月31日前 M Plan 的新订阅用户可享受5折优惠！！想体验该模型的，这波薅羊毛机会绝对不能错过！

[Vib3Coded @vib3coded]

PROMPT:

Build "Pixel Ice Rink" — a single-file, real-time three.js (r160, ES module via import map from jsDelivr) page: an isometric diorama of a floating outdoor ice rink, rendered as crisp pixel art in the style of "3D rendered as pixel art". UI text in English.

RENDER PIPELINE
- Orthographic camera at ~31° elevation, azimuth 45° + 90° steps (animated turns, drag sideways or buttons), wheel/pinch zoom. Fit the whole floating slab into the free screen area (clear of the title, a top-centre scoreboard and a right control rail).
- Render the scene into a tiny target (screen / pixel size, ~480×300 on desktop) with no anti-aliasing. Snap the camera so the world origin lands on a whole pixel (static things never shimmer).
- Post pass on a full-screen quad:
- depth-based outlines: darken a pixel that stands >0.55 m in front of a neighbour;
- map every pixel to the nearest colour of a fixed ~49-colour winter palette in OKLab (snow/ice greys, ice blues, reds, lamplight/wood/skin, pines, dusk peach/rose/mauve/lilac, brass/ochre);
- optional light 4×4 Bayer dither;
- upscale with texelFetch so each art pixel is an exact block of device pixels. The background (depth = 1) is the page colour.
- Hard sun shadows (BasicShadowMap, 2048). Lambert flat-shaded low-poly materials.
- Planar ice reflection: render the scene mirrored (scene.scale.y = -1 about the ice plane, ground, ice, particles and lamp posts hidden) into a second target. Sample it in the ice shader at gl_FragCoord. Gloss = small base + extra where the ice is wet, reduced where it is worn.
- A "Raw 3D" mode renders the same scene at full resolution with MSAA, for comparison.

SCENE
- A floating slab of snow over earth over rock, each extruded from a ragged rounded rectangle.
- A 24×12 m rink with 3 m corners: white boards with a yellow kick plate and red cap, built as band segments with gaps for two bench doors and a double resurfacer gate (hinged doors that swing open).
- Ice markings painted texel-by-texel into a 10 px/m canvas: red centre and goal lines, blue lines, circles, face-off dots, creases.
- Red-post goals with alpha-tested grid nets; players' benches; a lit scoreboard drawn with 3×5 pixel digits.
- A log warming hut with lit windows and chimney smoke; a garage for the resurfacer; lamp posts with point lights; string lights at night.
- Snowy stacked-cone pines, snowbanks, a snowman.

PLAYERS — procedural rigs of ~16 boxes
- Placed every frame from heading, lean, crouch and turn roll. Two-bone IK for the legs (skating stride with push-out and toe-out, gliding, sideways hockey stop) and for the arms holding a stick.
- Stick poses: carry/stickhandle, slap-shot wind-up and strike, pass sweep, celebration with stick raised.
- Goalies: pads, blocker and glove, crease positioning, butterfly drop on incoming shots.

GAME
- Two-on-two pick-up hockey plus goalies. Steering with arrival, separation and board/net avoidance; hockey stops when reversing hard.
- Carrier AI weaves to the slot, shoots when the lane is open, passes when pressured. Defenders pressure goal-side and poke-check; the nearest skater chases a loose puck.
- Puck physics with substeps: friction, board bounces, posts, solid nets with goal detection through the mouth, goalie saves (deflect or catch and play out), blocks by skates.
- Face-offs, goal celebrations, score, shots and a game clock on the scoreboard. Clicking the ice drops a puck.

ICE WEAR
- An RG DataTexture over the rink (20 px/m). Blades write scratches as they glide; stops gouge wider marks and throw spray particles that settle as wear; falling snow settles too.
- "Resurface the ice": players skate off through the bench doors; the resurfacer drives out of its garage through the gate and does shrinking rounded-rectangle laps (swinging in front of the nets), wiping its swath and leaving wet, reflective ice that dries over time. Then the players return for a face-off. Show an "ice worn %" bar.

UI
- Pixel display font (Silkscreen) with a clean body font.
- Light and dark themes via tokens; the background is a near-white or near-black page.
- Controls: Day / Dusk / Night (lighting presets blended over ~1.6 s), Snow Off / Light / Heavy, Resurface, Drop a puck, Pause, New game, Turn left / right, Pixel art / Raw 3D, Pixel size slider, Palette / Dither / Outlines toggles.
- A "How it's drawn" panel explaining the pipeline. Stacked mobile layout with no horizontal scroll.

[Vib3Coded @vib3coded]

$0.21 vs $7.88 for a tiny ice rink

MiniMax M3.1-Flash vs GPT-6 Astra

MiniMax takes this one for me. Look at the nighttime lighting: warm lamps, shadows across the ice, glowing windows, and all those little details. The whole scene feels more alive

And all that for just 21 cents - almost 38 times cheaper than Astra

I’m picking MiniMax for both quality and price. What about you?

[Liam Wilson @LiamWilson7787]

@VoxcatAI 这个太cool了，图片拿到混元里，直接建模

[VoxCat @VoxcatAI]

如果角色有一座只属于自己的微型住宅，会是什么样？

风格基调：日系生活方式杂志 × 微缩建筑设计 × 手工艺产品摄影 × 温柔幽默的角色叙事。

核心玩法：把角色的身份信息转译为建筑语言——道具决定轮廓，主色决定材料搭配，纹样进入门窗与五金。再用“角色从窗口探出”的尺度反差，让居所拥有生活感。暖白背景、真实材料和黑色大字，则负责统一整套封面的视觉秩序。

提示词：

【角色】，设计为一张中文版日系生活方式设计杂志封面。Japanese Lifestyle Magazine × Editorial Still Life Photography × Minimalist Object Design × Playful Architecture。

画面中央陈列一座专门为角色设计的微型居所，具有完整可居住空间、精致手工结构与独立设计产品的造型美感。以圆润几何建筑为基础，根据角色身份、标志性道具、所属作品、代表配色与文化符号自主设计建筑轮廓，将识别元素融入屋顶、外墙纹理、门窗、提手与装饰结构。

材料根据角色气质选择藤编、竹编、木材、织物、陶瓷、皮革或金属，形成协调的真实材质组合。窗口内可见少量精致家具、灯具与角色相关生活用品，传达温暖、完整的居住感。

角色安静待在居所内部，从正面窗口自然探出头部或上半身，目光看向镜头。动作根据人物气质自主设计，可托腮、扶住窗沿、捧杯或轻轻挥手，神态慵懒、好奇、从容，略带幽默感。人形角色呈现明确成年真人化的微缩形象，保留自然成年面部结构、真实皮肤��理与发丝；动物保留真实毛发与动物结构。清楚保留角色核心身份特征，形成角色与建筑之间可爱的尺度反差。

3:4竖版，平视略微俯拍的产品摄影视角。微型居所居中独立陈列，占画面中部约45%，主体轮廓完整，窗口与角色面部清晰，四周留出宽松的编辑排版空间。

纯净暖白色无缝摄影棚背景，背景与地面自然衔接，主体下方具有柔和真实的接触阴影。高调柔和漫射光，明暗过渡细腻，清楚呈现编织纹理、材料厚度、织物纤维、木纹、陶瓷釉面与手工制作痕迹。整体真实、温暖、干净，具有高品质家居产品摄影质感。

顶部约四分之一区域使用超大黑色连笔斜体中文刊头“生活与家”，形成舒展醒目的杂志标识。底部使用一行大号黑色现代中文标题，根据角色自动拟定，围绕“角色与家”展开。少量简体中文编辑短句、期号、细线与几何圆形标识分布于版面边缘，建立清晰的字号层级与现代编辑秩序。

整体配色根据角色身份自主协调，维持明亮中性背景、真实材料色彩与黑色字体的克制关系。日系生活美学、手工艺建筑、安静生活叙事与温柔幽默共同构成封面气质。

右侧约5%边缘位置保留极小竖排小写“voxcat”，末尾带一颗青金石蓝色圆点。

[Florence @flxrnc]

I made this 3D animation in a few hours with Opus 5.5 + Tripo + Blender

this is f*cking insane…

every single object in this video was made with Tripo

Check the PDF below to learn how to create the same with Tripo and Opus 5.5

Then read the article, it's everything you need to master motion design with Opus 5.5

[Viktor Oddy @viktoroddy]

❤️‍🔥 Unlimited ANIMATED prompts for AI websites from the video in one click: https://t.co/8ebpHNiVOM

[Liam Wilson @LiamWilson7787]

@viktoroddy very nice!

[Chong-U @chongdashu]

Here's a longer gameplay video of the my cosy low poly 3d game Retro del Sol.

It was made with Opus 5.5 and GPT 6 Astra - all assets by Retro Diffusion 3D and animated in Blender.

Fully multiplayer and online in your browser.

Works on mobile and desktop.

[あぎ @agi_aibusi]

【GPT Image 2.5でダンスモーション】

かわいいオリキャラの制作から
16コマの連続したダンスモーションまで
一撃で簡単に生成！

キャラの衣装、顔の一貫性を保ちながら
連続的な動きを生成でき
Gif化やショート動画に活用できます

そのまま使えるプロンプトはツリー⇩
※実践プロンプト30選は引用元の記事から

[岚叔 @LufzzLiz]

复刻差不多了，给大家分享教程并开源代码，先看看我的录屏效果如何，主要看前两个，后面还没做😂

首先特别感谢下原作者@DilumSanjaya，创造出如此美妙的项目，让我大受震撼👍

【手把手教程】
1️⃣ 直接将原作者视频发给coex，让其复刻视频的网站。codex 会截图、抽���、生图然后进行默默复刻

2️⃣ 网站复刻很简单，一般十几分钟就OK了，然后我们有了初级形态，细胞效果这时候会用Three.js 程序化 3D展示，但是要达到作者的效果Three.js 肯定是不够的。我尝试用GPT image 2 堆不同角度去实现，但是效果不佳，于是进行第三步

3️⃣图生3D，让GPT image 2 生成细胞精美的主视图、左/右视图、背面。然后找个可生成3D模型，这里我们用的是tripo。生成好后就可以下载下来给codex，让其换成真3D文件

项目开源见评论👇
ps：时间有限我们只做了前两个3D效果，后面好实用的three.js堆的，还没改，我们后续再去维护，我已经想到一些更好玩的，想要去尝试🐶

晚点我��补一个在线体验地址给大家～