
Meshy is an AI 3D model generator that turns text or images into 3D models in under a minute, with printability checks, auto-repair and direct export to slicers like Bambu Studio and OrcaSlicer, plus assets for games and media.
Emerging & Specialized Review
Zoo Design Studio is an AI-native CAD application for hardware design that combines familiar point-and-click modelling, code-based CAD and an AI engineering agent called Zookeeper, all built on Zoo's own geometry engine producing editable B-rep models rather than meshes.

Very good
Zoo earns 4.0 as one of the more serious attempts at AI for engineering: Zoo Design Studio produces editable B-rep CAD geometry, not meshes, and lets engineers mix point-and-click, code and the Zookeeper agent in one workspace on Zoo's own geometry engine. It is promising for hardware engineers and startups iterating quickly. The platform is still expanding, with analysis and manufacturing announced rather than shipped, so established CAD teams should treat it as a complement for now and verify AI-proposed parts. AInexfinder Editor Score: 4.0/5.
Best for: Mechanical part design; AI-assisted CAD
The visit link goes to zoo.dev, the official site. Paid placements on AInexfinder are always labelled. How we make money
Most AI 3D generators produce meshes that look good but are hard to edit or manufacture. Zoo takes the opposite approach: an AI-native CAD tool that creates precise, editable engineering geometry.
Zoo Design Studio combines familiar point-and-click CAD, code-based modelling and an AI engineering agent called Zookeeper, all on Zoo's own geometry engine producing B-rep models. Compared with AI model generators like Meshy, it is built for engineering parts rather than visual assets.
This review covers how Zoo Design Studio models a part, which engineers it suits, how it is priced, and where it cannot yet replace an established CAD suite.
| What it is | An AI-native CAD platform for hardware design |
|---|---|
| Modelling | Point-and-click, code-based CAD and AI prompts |
| Agent | Zookeeper engineering agent |
| Geometry | Editable B-rep models, not meshes |
| Engine | Zoo's own geometry engine |
| Roadmap | Design available; analyse and manufacture announced next |
| Platforms | Windows download and a browser version |
| Access | Free to try, with paid options |
Zoo's core proposition is that AI should work with real engineering geometry. Its Design Studio produces B-rep models, the precise boundary representations used in professional CAD, rather than meshes, so parts stay editable, dimensionable and closer to manufacturable.
It also lets engineers work however they prefer. You can sketch and extrude with familiar point-and-click tools, write the design as code, or ask the AI agent to propose and build geometry, and switch between those modes in one workspace.
The AI side is more than a generator. Zookeeper is described as an engineering agent that researches, reasons about constraints, plans and proposes designs, and helps prepare parts for manufacturing, working directly with the actual geometry rather than producing a picture.
The honest framing is that Zoo is an ambitious, technically interesting platform that is still expanding: design is available now, while analysis and manufacturing capabilities are announced as coming next. Teams with mature CAD workflows should evaluate it as a complement before a replacement.
Zoo produces editable B-rep engineering geometry, not meshes, which is what makes it more than another AI 3D generator.
Zoo Design Studio offers three ways to model, on top of its own geometry engine.
You can model with familiar CAD tools, define parts in code for repeatable, parametric designs, or describe intent in a prompt and let the agent build. Each route produces the same editable geometry, so a prompt-generated part can be refined by hand or in code.
For organic or visual 3D assets rather than engineering parts, generators such as Sloyd or Rodin are better suited; Zoo targets mechanical design.
Zookeeper works with the model itself: it can research, reason about constraints, propose changes and evaluate manufacturability as you design. Underneath, Zoo's own geometry engine creates B-rep models, and its 'Aquarium' gallery shows real, editable models built in the software, from gears to turbine blades.
Because the agent can propose geometry quickly, engineers should verify dimensions, tolerances and constraints before relying on a part, exactly as they would review a colleague's work.
Zoo suits mechanical and hardware engineers, product designers and makers who want precise, editable CAD with an AI assistant, and teams interested in code-based, version-controllable design.
It suits startups and individual engineers exploring how AI can speed up early design iterations, and engineers who like scripting parametric parts rather than clicking through every feature by hand.
It is less suited, for now, to organisations deeply invested in an established CAD suite with years of models, plugins, simulation and PDM workflows, where switching costs are high and analysis and manufacturing integration matter.
A practical approach is to use Zoo on new parts or side projects, judge the agent's usefulness on real design problems, and track the platform's analysis and manufacturing features as they arrive.
Zoo Design Studio can be tried for free, with paid options for heavier use and teams, and an enterprise offering. We do not publish prices, which change; check the site.
The value question is design speed and quality: if Zookeeper and code-based modelling cut iteration time on real parts, the cost can be justified, especially alongside existing tools.
The sensible approach is to model a real part you know well, compare the time and result against your usual CAD workflow, and decide based on that rather than demos.
Remember to include the cost of learning a new tool and of recreating any templates, standard parts or design conventions you rely on in your current CAD system, which can outweigh the subscription itself for an established team.
Because engineering parts must be right, test Zoo on a part with known requirements.
AI-native CAD
Point-and-click modelling
Code-based CAD
Zookeeper AI agent
B-rep geometry engine
Manufacturability checks
Browser and desktop versions
Example model gallery
The useful comparison is with other 3D tools, by whether they produce engineering geometry.
| Tool | Tool | Output | Best for |
|---|---|---|---|
| Zoo | Editable B-rep CAD | AI-native engineering design | Mechanical and hardware parts |
| Meshy | Meshes from text or images | Printable 3D models | Makers and game assets |
| Sloyd | Parametric 3D assets | Game-ready models | Game asset pipelines |
| Rodin | AI 3D models | Detailed visual models | 3D creators |
Last checked . Read from zoo.dev during this check, with the screenshot above captured at the same time.
Zoo is still expanding: analysis and manufacturing features are announced rather than shipped. Engineers should verify AI-proposed geometry, dimensions and tolerances before relying on parts.
Three AI 3D generators to compare with Zoo's engineering-focused CAD.
AInexfinder does not list plan prices or billing details. For current pricing, plans, and trials, visit the official Zoo website.
Visit official website for pricingVendor pricing, credits, and billing policies change over time. Always confirm on the official site before you buy.
Zoo earns an AInexfinder Editor Score of 4.0/5 (very good). Best for: Mechanical part design; AI-assisted CAD; Parametric code-based design.
Visit ZooLog in to write a review.
No reviews yet. Be the first to share your experience with Zoo.
Assigned reviewer
Daniel ReedSenior AI Tools Reviewer
Daniel leads AInexfinder's tool reviews. He researches each tool from its own pricing, documentation and policies, checks its claims against independent sources, and writes up what it does well, where it falls short and who it genuinely suits.
Daniel and the AInexfinder editorial team research Zoo from its own website, pricing, documentation and policies, checked against independent sources. This review is research-based: unless its method note says otherwise, we have not bought a plan or run hands-on tests. The Editor Score is our editorial judgement, and payment never changes it.
Related Pages