SKY-001: Skyfall AI's 'Engineering World Model' for CAD/BIM
Skyfall AI (co-founder/CEO Sam Pasupalak, ex-Maluuba) unveiled SKY-001 on Oct 19 on X: a spatial world model that plans engineering and architectural designs directly in the visual world via a two-stage Hierarchical Composer + Construction World Model, covering CAD, BIM, PCB, aerospace and molecule design. Access is waitlist-only.
Why it matters
Plans designs in stages — overall structure, then assemblies, then part geometry — instead of generating everything in one shot, so projects with hundreds or thousands of steps stay consistent. Company claims: 51-125x faster, 150-210x cheaper ($0.005 vs $0.75 per CAD sample), and 79.7% vs 79.1% IoU against GPT-6 Astra. All figures are Skyfall AI's own tests and have not been independently verified.
What you could build with it
An indie developer building a low-cost design co-pilot for small architecture studios could pair SKY-001's staged planning with a sketch-to-bill-of-materials pipeline, turning rough floor plans into priced component lists and verifying whether the staged approach really keeps thousand-step projects consistent.
Does it hold up?
Too early to judge — the model is waitlist-only and every performance figure is Skyfall AI's own test; Business Today and Economic Times both note the numbers have not been independently verified.
Built with SKY-001
- Skyfall AI says SKY-001 beats frontier models on CAD and BIM (RuntimeWire)runtimewire · Oct 20 writeup of the Oct 19 X thread: staged visual planning, claimed speed/cost gains, waitlist access, and the caveat that benchmark details remain limited.
- Indian AI engineer builds 'world model' (Economic Times)economictimes · Coverage of Pasupalak's background (Maluuba, acquired by Microsoft) and the model's 2D-image/text-to-3D-assets approach; notes the figures come from the company's own tests.
Learn more
press coverage of the launch (Business Today) →
First spotted on x: source.
More AI models
OpenAI publishes 722 AI-generated math manuscripts from an unreleased modelOpenAI released 722 mathematical manuscripts in 372 result families to GitHub under Apache-2.0, produced by an…models · JEV 0.74Tencent Hunyuan Hy3: open 295B MoE with 21B active under Apache 2.0Tencent open-sourced Hy3, a 295B-parameter mixture-of-experts model with only 21B parameters active per token, under…models · JEV 0.73NVIDIA Isaac Lab 3.0 GA + Isaac OS open runtime + Agile ONE SN humanoidMajor Isaac platform expansion at GTC Berlin Oct 21: Isaac Lab 3.0 is generally available as the open-source robot…models · JEV 0.73StepFun Step-5-Preview — open BF16 weights releasedStepFun published the full BF16 weights for its flagship Step-5-Preview agentic model (600B-parameter sparse MoE, 27B…models · JEV 0.73NVIDIA Nemotron 3.5 Retriever collection: Embed 8B/1B, Rerank 8B, Parse 631MNVIDIA's Oct 21 collection drop for agentic retrieval: Nemotron-3.5-Embed in 8B BF16, 1B BF16 and 1B NVFP4…models · JEV 0.72Kimi K3.1 Pro: Moonshot's 2.8T open-weight flagship officially launchesMoonshot AI officially launched Kimi K3.1 Pro on Oct 23, a mixture-of-experts model with 2.8 trillion total / 104…models · JEV 0.72
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.