Tag
Someone attempted to push Claude Opus 5.5 to its limits using Houdini and concluded there is no ceiling in sight — the entire scene, including final color grading, was completed via iPhone voice commands with MCP, and rendered for 50 hours on an RTX 5090, showcasing the enormous potential of AI-driven creative workflows.
Someone gave Opus 5.5 a single task: to build an ancient Greek world as detailed as possible. Everything visible in the scene is generated entirely by code, with no pre-made assets used.
@Saccc_c thanks supporters of the Little Prince game project, reporting extremely high website traffic and announcing that the game now supports eight languages (Chinese, English, French, Japanese, etc.), while inviting bug reports and screenshots for later fixes using Grok. The project, built in 16 days using Opus 5.5, Tripo and Blender as a 3D story game, is planned to be fully open-sourced along with a detailed creation walkthrough.
一位开发者用 Rust 从零编写了自定义 3D 物理引擎和 WebGL2 渲染器，通过 WebAssembly 在浏览器中运行，不依赖任何外部数学或物理库，约 1200 行代码，.wasm 二进制小于 100KB 且运行在 120 FPS。
LittleJS 发布了自推出以来的最大更新，新增完整的 3D 渲染器、内置关卡编辑器、粒子效果库、性能大幅提升，并加入了面向 AI 编码助手的工具包和 Claude Code 插件。
Bilawal Sidhu recreated San Francisco's Fleet Week as a photorealistic 3D digital twin, using real flight and vessel tracking data to drive camera tracking and speed-ramped jet flyovers.
A showcase of camera-panning in a 3D fluid-flow-around-a-sphere simulation, computed via the finite element method on the Navier-Stokes equations. The tweet demos an interactive/animated CFD visualization tool.
This tweet showcases an interactive 3D simulation of Apollo 11 Mission Control created using Opus 5.5 AI, featuring real NASA audio, synchronized subtitles, and interactive engine elements.
This appears to be a social media post about integrating GPT-6 Sol and Opus 5.5 AI models with Unreal Engine and Blender software, likely for applications in 3D rendering or AI development.
A new checkpoint for Gemini 4 Pro, codenamed barium-b, has appeared in Arena, showing noticeable improvements in 3D F1 studio output with better detail, materials, and UI coherence. The update is part of Gemini 4's post-training phase, with Google aiming for a fast release while safety work continues.
OmniDICOM is a cross-platform software for viewing and editing DICOM medical images, featuring metadata editing, 3D rendering, and PACS/DICOMweb integration, available for Mac and Windows with a subscription model.
A user shares their experience using Remotion for creating videos, highlighting issues with 3D rendering and suggesting improvements for the preview interface.
This article provides a practical guide on using Three.js's native Gaussian Splatting support, explaining how to load and render Gaussian Splats in web 3D scenes with code examples.
A Twitter user discusses the performance of Fable 5.1 in Blender, comparing its output to a reference image and noting it is good but not perfect.
A tweet highlighting the potential of WebGPU for building impressive interactive demos, featuring a playful challenge about a cube falling into a hole.
Orbify demonstrates its curved mapping technology for turn-by-turn navigation with interactive 3D controls, featuring patent-pending warping technology.
This article compares the performance of three AI models—Muse Spark 1.2, Claude Opus 5, and Kimi K3—in generating code for a 3D shopping mall scene, highlighting trade-offs in cost, speed, and code quality.
A tweet showing a trick of handing a floor plan to AI and having it directly generate a 3D render in a specified style (e.g., Kyoto style), reflecting the power of current 3D generation models.
Andrej Karpathy experiments with Opus 5, giving it the first paragraph of Lord of the Rings and a 1M token budget to create a 3D JS rendering of the story, highlighting LLM stamina for hyper-custom worlds while noting weaknesses in multimodal auditing and gameplay.
Andrej Karpathy shares an experiment where Claude Opus 5 used a 1M-token budget to procedurally render Lord of the Rings in three.js, sparking thoughts on LLM testing, custom world generation, and multimodal weaknesses.