AI MBSE
AI for SysML v2
MBEstudio is an AI-native MBSE studio. Describe a system in plain language and its agent writes a SysML v2 model, checks it with the compiler and repairs what fails. Then every requirement is evaluated against the model, with its value, limit and margin.
What the AI does
From a description to an evaluated SysML v2 model
Generates the model
The agent turns a system description into SysML v2 parts, ports, connections, equations and requirements, validates it through the compiler and repairs what fails.
Maps your existing data
Imports map KiCad netlists, Onshape BOM CSV, ReqIF, MATLAB and CSV tables onto existing parts, so the AI works on your design rather than a blank page.
Works from your own agent
Claude Code, Codex, Cursor, Claude Desktop and ChatGPT connect over MCP to validate, evaluate, edit and simulate models with the same compiler the studio uses.
Why SysML v2
A text model an AI can write and a compiler can check
SysML v2 has a textual syntax, so a system model can be written, compared and reviewed like source code. That is what makes it a practical target for AI: the model is text, and the checks are deterministic.
MBEstudio keeps the SysML v2 text as the single source of truth. A change in the 3D view, a diagram or a table becomes an edit to the text. The studio, the MCP server and the command line run the same Rust compiler, built to WebAssembly, so the verdict in the browser is the verdict an agent or a CI job gets.
Every evaluated value says whether it was stated, computed from an expression or estimated and flagged for review. A missing input stays indeterminate and never defaults to a pass, so an AI-generated model cannot quietly look finished. Where MBEstudio fits against CAD and traditional MBSE tools.
What you can check
Engineering evidence, not just a diagram
Requirements across states
Run the declared state machines and see each requirement value, limit and margin over time, with the intervals where it applies.
Design trades
Sweep design variables and alternatives against declared objectives and constraints, compare the Pareto front and apply a candidate to the model.
Interfaces, budgets and geometry
Follow typed ports through internal block diagrams and 3D, sum mass and power budgets, and export SysML, STL, CSV and HTML reports.
Tool trade
Where MBEstudio sits against the tools you already use
| Tool | What it answers | What MBEstudio adds |
|---|---|---|
| Desktop MBSE suites (Cameo, Capella) | What are the functions, states and interfaces, drawn and managed at enterprise scale? | A SysML v2 text model in the browser with nothing to install, checked by one compiler that runs the same in the studio, in your agent and in CI. |
| AI-first web MBSE tools | Can a model be drafted from a document or a chat? | Every AI edit passes the compiler and is repaired before you see it, every value says where it came from, and a missing input stays indeterminate instead of passing. |
| CAD (CATIA, SolidWorks, Onshape) | Will it manufacture? Will it assemble? | Puts the architecture in metres and kilograms earlier, with geometry, mass and power derived from the same model text. |
| Requirements tools (DOORS, Jama) | What did we say we would do? | Evaluates each requirement against the model with its value, limit and margin, across the states where it applies. |
| Spreadsheets and scripts | What is the roll-up? Which option is best? | The same roll-ups and Pareto trades computed from the model, so they cannot silently disagree with it. |
By name
Thirteen SysML v2 tools, side by side
Flow Engineering, SysGit, Dalus, Eclipse SysON, Innoslate, Syside, Ansys SAM, Visual Paradigm, SysModeler, Valispace, Siemens and CATIA Magic / Cameo, with where each runs, who it suits best and how MBEstudio differs. See the full SysML v2 tool comparison.
Deep dive
MBEstudio against traditional MBSE
Desktop suites such as Cameo are mature, widely adopted and now edit SysML v2 as text and simulate it. The differences are in the architecture.
The text is the model
The model is a plain .sysml file you can diff, review and keep in git. An edit in 3D, a diagram or a table becomes an edit to that text, so there is one source of truth.
One compiler everywhere
The studio runs in a browser with nothing to install. The same Rust compiler, built to WebAssembly, gives the verdict in the studio, the MCP server, the command line and CI.
Metres and kilograms
Parts carry geometry, mass and power from the model itself, so budgets, clearances and requirement margins are evaluated continuously rather than at review.
Use it from your agent
One MCP server for AI SysML work
Add MBEstudio to your agent with its MCP URL, sign in with your account email, and ask for an estimate before any paid job starts. Setup steps for each agent.
https://studio.vectorstreamsystems.com/mcp
FAQ
AI and MBSE questions
What is AI MBSE?
AI MBSE is model-based systems engineering where an AI model helps write, check and change the system model. In MBEstudio the model is SysML v2 text, so an AI can author it the way it writes code, and a compiler checks every result before you rely on it.
Can AI write a SysML v2 model?
Yes. Describe a system in plain language and the MBEstudio agent generates a SysML v2 model, validates it through the compiler and repairs what fails. You review the parts, ports, connections and requirements it declared in the studio.
How do I know the AI's model is right?
Every value says whether it was stated, computed from an expression or estimated and flagged for review. A missing input stays indeterminate and never defaults to a pass, and requirements show their calculated value, limit and margin.
What is an AI SysML v2 tool?
An AI SysML v2 tool writes and edits system models in the SysML v2 textual notation from plain language. MBEstudio's agent generates SysML v2, runs it through a compiler and repairs what fails, so every model you open has passed the same checks as one written by hand.
How does MBEstudio compare to Cameo?
Cameo is a mature desktop MBSE suite that now edits SysML v2 as text and simulates it. MBEstudio runs in a browser with nothing to install, keeps the .sysml text as the single source of truth, checks every AI edit with the same compiler in the studio, MCP server, command line and CI, and evaluates geometry, mass, power and requirement margins from the model.
Which AI agents can use MBEstudio?
Any agent that speaks MCP. Claude Code, Codex, Cursor, Claude Desktop and ChatGPT connect to https://studio.vectorstreamsystems.com/mcp and use the same compiler to validate, evaluate, edit and simulate models.
What does it cost?
The Individual plan is $39 a month and includes 2,000 AI credits a month. Credits cost $0.01 each and can also be bought as one-time packs without a subscription. Every AI job shows its maximum cost before it starts.