---

id: knowledge-concept-2026-04-15-can-frontier-llms-solve-cad-tasks
title: Can frontier LLMs solve CAD tasks?
type: concept
status: draft
confidence: medium
sources:
- raw/articles/2026-04-15-can-frontier-llms-solve-cad-tasks.md
backlinks: []
last_compiled_at: "2026-04-15T10:04:33Z"
last_verified_at: "2026-04-15T10:04:33Z"
owner_agent: hermes

---

Can frontier LLMs solve CAD tasks?

summary

Can frontier LLMs solve CAD tasks? Can frontier LLMs solve CAD tasks? Kerrick Staley About Can frontier LLMs solve CAD tasks? Feb 22, 2026 Frontier LLMs like GPT-5.3-Codex, Gemini 3.1 Pro, and Claude Opus 4.6 have spiky capabilities, performing several stddevs above median human performance at some tasks while failing at some “easy” ones. LLM pretraining datamixes often emphasize general knowledge, reasoning, and coding. The human “training mix” includes far more samples of visual/spatial/motor tasks, which come about naturally in the embodied human experience. World models like Sora and Genie are pretrained on video and 3D video game data and excel at predicting the behavior of the real world. But no current model is at the frontier of both reasoning/coding and spatial reasoning/world modeling. We’d expect (and it seems empirically true) that LLMs trained primarily on text are worse...

source coverage

• primary source: raw/articles/2026-04-15-can-frontier-llms-solve-cad-tasks.md