IA Lab
AI experiments, AI+Art crossovers, and curated external articles about artificial intelligence and creative technology.
NVIDIA Lyra 2.0 — Give It One Photo, It Gives You an Infinite Explorable 3D World
NVIDIA’s Spatial Intelligence Lab just dropped Lyra 2.0 — open source, published April 2026 — and it…
Tripo AI H3.1 — Native 3D Diffusion Just Made Film-Grade Assets in 2 Seconds
One month ago, Tripo gave us Smart Mesh P1.0 — the fastest game-ready low-poly generator around. Now…
HappyHorse 1.0 — Alibaba’s Anonymous #1 Video Model Is Open Source and It Just Beat Everyone
On April 7, 2026, a model with no readme, no team page, and zero launch hype appeared…
Kling 3.0 — Sora Is Dead, Long Live the AI Video King
OpenAI is shutting down Sora on April 26, 2026 — that’s six days from now. The platform…
Meshy × Formlabs — From Text Prompt to 3D-Printed Object in Under 5 Minutes
You type a prompt. Thirty seconds later you have a 3D model. Five minutes after that, it’s…
Roblox Studio Goes Agentic — Type a Prompt, Get a Fully Tested 3D Game
Roblox just flipped the game development model upside down. Not “AI-assisted” in the polite, copilot sense —…
Adobe Firefly AI Assistant — One Prompt to Rule All Your Creative Cloud Apps
Adobe just turned your Creative Cloud subscription into something it was never supposed to be: an AI…
HunyuanWorld 2.0 — One Photo In, Full 3D World Out. And Tencent Just Open-Sourced It
One photo. A forest clearing, a sci-fi corridor, a medieval courtyard. That’s all HunyuanWorld 2.0 needs to…
Framestore’s Superman — 4D Gaussian Splatting Just Conquered Hollywood
Forty shots. Two actors. One 192-camera volume. No traditional renderer. Framestore just shipped the first use of…
GaussianGPT — Forget Diffusion: TU Munich Just Taught a Transformer to Build 3D Scenes Token by Token
Every 3D generation model you know uses diffusion. FLUX, Stable Diffusion 3D, Shap-E, Point-E — they all…