AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
Multi-Agent/
Multi-Agent-CAD
Multi-Agent-CAD logo

Multi-Agent-CAD

Active·★ 1.0k·MIT·Updated 2026-10-17
★ Essential★ Trending

MAC (Multi-Agent CAD) turns a natural-language design brief into editable engineering geometry using build123d. It can generate a single printable part, including print-in-place mechanisms, or decompose a mechanism into separately generated components and assemble them. Outputs are STEP, STL and GLB rather than render-only results, and assemblies can be exported as URDF for robotics simulation.

v2 adds visual verification and an independent Judge to the original single-part workflow. On the project's documented ten-prompt, 141-feature benchmark, both MAC v1 and v2 passed 140/141 features while using far fewer tokens than a reproduced CAD-skill baseline (for v2, 1.40M tokens and about $4.85). The README is careful to say these figures do not apply to assembly generation, which is a technology preview with curated examples.

Setup is a conda environment plus a pinned aider-chat install, with build123d held below 0.13 because of the bundled OCP 7 runtime. The default provider is an OpenAI-compatible DashScope (Qwen) endpoint. There is a web UI. License is MIT.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool. Data as of 2026-10-22.

License
MIT
Latest release
v1.0.0 · 2026-09-24
Last commit
2026-10-17
Contributors
2
Forks
98
Open issues and PRs
1
First published
2026-07-30

Source: GitHub, npm and PyPI public APIs.

#3d-modeling#3d-printing#agent-framework#agent-pipeline#aider#assembly#build123d#cad
$ Install
$ git clone https://github.com/Pan-Chera/Multi-Agent-CAD.git && cd Multi-Agent-CAD && conda env create -f environment.yml && conda activate multi_agent_cad && pip install --no-deps "aider-chat==0.82.3"
↗ Visit site★ GitHub
01

Features

01Natural-language input for geometry, dimensions, interfaces and motion
02Single-part workflow scoring 140/141 features on its documented ten-prompt benchmark
03Visual plus geometric verification with an independent Judge in v2
04Experimental multi-part assembly generation with URDF export for simulation
02

Why choose it

+Produces editable STEP/STL/GLB geometry, not just images, so results can go to slicers and CAD tools.
+Strong single-part benchmark (140/141 features) at a fraction of the token cost of a skill-based baseline.
+URDF export connects generated assemblies to robotics simulation.
03

Trade-offs

!Assembly generation is an experimental preview; geometry and joints must be validated before manufacturing.
!Dependency setup is fiddly, with NumPy conflicts between aider-chat and build123d and a pinned build123d version.
!Large assemblies can consume millions of tokens per generation.
04

Maintenance check

ActiveChecked: 2026-10-20
Last commit
2026-10-17
Commits, last 90 days
67
Releases, last 90 days
2
Latest release
v1.0.0 (2026-09-24)
Open issues
1
Recent issues with at least one comment
6/6

Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.

05

Compatibility

build123d
CAD kernel
Verified via docs
DashScope (Qwen)
Default provider
Verified via docs
OpenAI / Anthropic
Providers
Verified via docs
URDF
Simulation export
Verified via docs
06

Quick start

1
$ git clone https://github.com/Pan-Chera/Multi-Agent-CAD.git
2
$ cd Multi-Agent-CAD
3
$ conda env create -f environment.yml
4
$ conda activate multi_agent_cad
5
$ pip install --no-deps "aider-chat==0.82.3"
07

Use cases

↳Generating 3D-printable parts and print-in-place mechanisms from a text brief
↳Prototyping robot mechanisms and exporting them to physics simulation
↳Producing editable STEP files for downstream CAD inspection
08

Alternatives

⚡Top VS:
VSagentscope-java★6.1k→VSMetaGPT★71.0k→VSdeer-flow★84.3k→VSMicrosoft AutoGen★61.5k→
agentscope-java logo
agentscope-java★ 6.1k
Build distributed, production-grade, long-running agents.
VSCompare→
MetaGPT logo
MetaGPT★ 71.0k
🌟 The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming
VSCompare→
deer-flow logo
deer-flow★ 84.3k
DeerFlow is a community-driven Deep Research framework, combining language models with tools like web search, crawling, and Python execution, while contributing back to the open-source community.
VSCompare→
Microsoft AutoGen logo
Microsoft AutoGen★ 61.5k
A framework that enables the development of LLM applications using multiple agents that can converse with each other to solve tasks.
VSCompare→
CrewAI logo
CrewAI★ 60.0k
Framework for orchestrating role-playing, autonomous AI agents. By working together, your Crew can tackle complex tasks.
VSCompare→
agentscope logo
agentscope★ 33.7k
AgentScope: Agent-Oriented Programming for Building LLM Applications
VSCompare→
adk-python logo
adk-python★ 22.0k
An open-source, code-first Python toolkit for building, evaluating, and deploying sophisticated AI agents with flexibility and control.
VSCompare→
Pydantic AI logo
Pydantic AI★ 20.8k
GenAI Agent Framework, the Pydantic way. A Python agent framework designed for production with type safety at its core.
VSCompare→
See all alternatives →

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs04Maintenance check05Compatibility06Quick start07Use cases08Alternatives
Pan-Chera /
Multi-Agent-CAD
↗

Text-to-CAD agents that output editable STEP parts and URDF-ready assemblies

1.0k98MIT
Last commit6d ago
StatusActive
CategoryMulti-Agent
Contributors2
Latest releasev1.0.0
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.