Fusion, not integration.
MathJet is a fusion of spreadsheet, scripting, and interactive visualization technologies — built from the ground up on a single C++ codebase as a modern, cross-platform alternative to MATLAB® and a more capable analytical companion to Excel.
Where most data analysis tools force a choice between paradigms —
spreadsheet or code, GUI or script, one language or another — MathJet
dissolves those boundaries. The same data shows up as cells, as
variables, and as plot elements; edit any view and the others update.
Python, R, .m files, and Excel formulas share one runtime
and one memory space. Open existing Jupyter notebooks and they run in
MathJet's own kernel with live variable inspection and interactive plots.
If you've felt notebook state hell — the cell you ran three hours ago
that made the rest of the analysis possible and now you can't remember
how — MathJet is for you. If you've watched a kernel restart eat thirty
minutes of loaded data, or caught an AI-generated matplotlib snippet
failing silently on your actual dataset, MathJet is for you.
Built for scientists, engineers, quantitative researchers, and data analysts doing rigorous multi-step work. Academic use is free.
A full Excel-compatible program implementing nearly all Excel
features — formulas, pivot tables, charts, conditional formatting —
and loading .xlsx files natively. First-class citizen
of the environment, not a data viewer or an add-on. Cells are also
first-class objects in the scripting runtime; reading a range from
Python is a memory access, not an I/O call.
Because the spreadsheet's formula engine and the scripting interpreters are the same code, you can write Python, R, or MATLAB function calls directly in cell formulas, alongside standard Excel functions. Formulas can also reference variables that scripts created earlier in the same session. Not a "UDF bridge," not an Excel add-in — the same interpreter.
All three share the same objects in memory, and MathJet's own interpreter is MATLAB-syntax compatible. A NumPy array built in Python can be read directly by R or by a MATLAB function call, with no conversion step in between. No kernel restarts, no half-remembered state from an earlier cell.
Native C++ rendering, not a web-canvas wrapper. Includes graphical value editing, axis folding, and interactive data grouping — patented capabilities that aren't available in any other tool. Click a plot; the underlying data highlights and the code that produced it jumps into view.
Open your existing .ipynb files directly. Run them
on a standard Jupyter or Spyder kernel exactly as today, or
switch to MathJet's kernel and the same notebook gains live
variable inspection without print(), editable
spreadsheet views of any DataFrame, and matplotlib
plots that become native interactive graphics — all without
changing a line of code.
Change a cell → plots update. Change a variable in Python → the R session sees it. Resize a range → the formula graph propagates. Everything is watched, everything is connected.
A pipeline view that makes multi-step analyses reproducible by
construction, and a native MCP interface so an AI assistant can drive
MathJet directly — inspect your spreadsheet, query your plots, modify
variables in place — instead of generating disconnected matplotlib code
that may or may not run on your data.
MathJet is the result of over ten years of development and is protected by five pending patents and a registered trademark. It's built by Jet Computing, LLC, an independent software company based in Newton, Massachusetts.
MathJet