OpenAI and Anthropic simultaneously launched GPT-6 Sol and Claude Opus 5.5, offering improved coding benchmarks, reduced API pricing, enhanced caching, and better long-task efficiency, with early demos showing impressive 3D scene generation capabilities.
The article details how Yingmu Technology’s cuNRTO paper, nominated for the RSS 2026 Outstanding Paper Award, moves nonlinear robust trajectory optimization onto GPUs, delivering up to 139.6× speedup while preserving 100% safety constraints, and situates this breakthrough within the company’s broader 3D‑to‑embodied‑AI research roadmap.
World Labs released three technically detailed papers—World Tracing, Modality Forcing, and Flex4DHuman—each extending 2D diffusion models to 3D generation, while co‑founder Christoph Lassner announced his departure due to injury, marking a notable milestone for the spatial‑AI startup.
PhysForge, a physics‑grounded 3D asset generation framework accepted at ICML 2026, converts a single input image into a fully interactive 3D object by first planning a hierarchical physical blueprint with a vision‑language model and then refining geometry, texture, and precise kinematic parameters via a diffusion model, supported by the large‑scale PhysDB dataset.
PhysX-Omni introduces a unified framework that models rigid, deformable, and articulated objects with full physical attributes, builds the large‑scale PhysXVerse dataset and PhysX‑Bench benchmark, and demonstrates superior performance over prior simulation‑ready 3D generation methods across geometry, scale, material, affordance, kinematics and description metrics.
CMU’s new ICLR‑2026 paper demonstrates how AI can move beyond static 3D model generation to create interactive scenes by learning both geometry and functional properties, enabling objects like doors and drawers to be manipulated, a step toward usable simulators for robotics and VR.
This week’s tech roundup covers the launch of Kimi K2.6 with a 300‑agent swarm capability and major performance gains, DeepSeek V4’s new sparse‑attention architecture and pricing, Meshy’s AI‑3D partnership, a $4.55 B AI‑brain funding round, Honor’s record‑breaking robot, M‑Flow’s cone‑graph memory engine, and Vision Banana’s unified visual model, all backed by benchmark data and industry commentary.
The article analyzes the fundamental challenges of 3D generative AI, compares existing NeRF, voxel, and single‑view methods, and explains how Microsoft’s open‑source TRELLIS.2 uses a field‑free O‑Voxel representation and 16× compression to produce 512³ full‑texture assets in about three seconds, with a step‑by‑step HyperAI demo.
World Labs unveiled RTFM, a real‑time frame model that runs on a single H100 GPU, generating persistent, interactive 3D worlds from 2D images without explicit 3D representations, highlighting the growing computational demands of generative world models and their potential to reshape AI-driven spatial intelligence.
This roundup highlights ten recent AI developments, from Apple's Matrix3D model that creates 3D scenes from photos, to Qwen's Deep Research assistant, Tencent's CodeBuddy 3.0, ByteDance's Seed1.5‑VL, Step Star's open‑source Step1X‑3D, Google's iOS icon refresh, Apple's eye‑tracking scrolling test, Chrome's upcoming Gemini AI assistant, Shanghai's AI Identity Ecosystem Alliance, and Kuaishou's Keling AI 2.0 topping the global video‑generation leaderboard.
The paper introduces G3PT, a groundbreaking autoregressive 3D generation model that employs a Cross‑Scale Querying Transformer and multi‑scale tokenization to produce high‑quality meshes from a single image, outperforming diffusion‑based methods and revealing a scaling law for 3D generation.
WonderTurbo introduces a real‑time 3D scene generation pipeline that accelerates both geometry and appearance modeling to under a second per view, using StepSplat, QuickDepth, and FastPaint modules, achieving up to 15× speedup while maintaining high visual quality.
This issue reviews the latest AI breakthroughs—including OpenAI’s O3 and o1 models, pricing cuts, new features in ChatGPT, product launches like Pika 2.0 and Gemini 2.0, a heated debate on pre‑training data bottlenecks sparked by Ilya Sutskever, a novel black‑box forgetting method, and DeepMind’s Genie 2 3D world generator—highlighting how industry dynamics and research directions are reshaping the field.
This weekly roundup highlights major AI developments such as World Labs' 3D world generation model and OpenAI's full‑scale o1 model, introduces new AI image‑generation features for WeChat articles, reports on upcoming Nintendo Switch 2, and shares industry insights, expert opinions, and miscellaneous tech news.
Tencent has open‑sourced its 389‑billion‑parameter Hunyuan Large Mixture‑of‑Experts model—featuring 52 B active parameters, 256 K token context, novel routing, KV‑cache compression, and advanced training optimizations that beat leading open‑source models—and its first text‑to‑3D/image‑to‑3D Hunyuan 3D Generation model, both downloadable via GitHub, Hugging Face, and Tencent Cloud.
This guide walks you through using Midjourney to generate a reference image, preparing it for AI processing, creating a 3D model with Tripo AI, refining it in Cinema 4D, and rendering a complete scene, highlighting both the workflow and current limitations of AI‑driven 3D modeling.
Kwai (Kuaishou) has eight papers accepted at CVPR 2024 covering multi‑dimensional human preference for text‑to‑image generation, short‑video quality assessment, efficient video quality assessment, compressed video enhancement, conditional unsigned distance fields, universal cross‑domain retrieval, perception‑oriented frame interpolation, and test‑time energy adaptation.
KuaiShou presents 8 papers at CVPR 2024, covering AI advancements in computer vision, video quality assessment, and 3D generation, showcasing cutting-edge research in machine learning and multimedia technologies.
This article explores the rise of AIGC-driven 3D model creation for e‑commerce, reviewing foundational technologies like NeRF, Point‑E, Shap‑E, DreamFusion and Magic3D, detailing practical experiments, hardware requirements, current challenges, and future industry implications.
