Jenner runs your existing SAS programs — DATA steps, PROC SQL, macros — without a SAS license, on Windows, macOS, and Linux. Same syntax, no per-seat fees, no rewrite.
SAS prices licensing per seat and per year, tracks CPU core counts, and requires annual renewal. Jenner has no per-seat license and runs as a single self-contained binary.
Jenner has no per-seat license and no annual renewal to re-justify. The same binary runs for one analyst or for fifty, so adding a person to the team is not a purchasing decision, and the cost of your analytics stack stops scaling with headcount.
Jenner runs natively on macOS — including Apple silicon — as well as Windows and Linux, all from the same source. Analysts work on the machine their employer issued them, without a virtual machine, a second operating system, or a remote session between them and their own code. The same program moves unchanged from a laptop to a Linux server.
Running SAS programs on a Mac →Decades of institutional analysis sit in SAS7BDAT and XPT files, and that is usually the part of a migration people expect to be hard. Jenner reads and writes both formats directly, alongside CSV and Parquet, so nothing needs converting before you start and results can go back out in the same format a colleague expects.
Replacing SAS is usually imagined as a rewrite, which is why it gets postponed for years. Moving to Jenner is not a rewrite — the programs stay in SAS syntax and keep their DATA steps, macros, and PROC calls. What follows is the sequence teams actually use, and most of them are running real work within a day.
Start with a program you know well and whose output you can recognize at a glance. There is nothing to port and no project to staff — install the binary, run the .sas file, and read the log. DATA step and PROC code runs as written, so the first run is the whole evaluation: you see your own output, from your own code, on the first afternoon.
Run the same program under SAS and under Jenner and compare the resulting datasets. PROC COMPARE exists in both and is the natural instrument: it reports observations, variables, and values down to a tolerance you choose. Writing to SAS7BDAT from both sides means the comparison is on identical ground, so what you get is a clean, reviewable result rather than an impression.
Accuracy is something you should be able to demonstrate, not just assert. Jenner passes all 58 datasets in the NIST Statistical Reference Datasets benchmark — the standard third-party accuracy suite for statistical software — and cross-validates its statistical procedures against the Python (statsmodels) and R implementations. That, plus your own PROC COMPARE output, is the package a reviewer needs to sign off.
Nothing requires a cutover. Jenner runs as one self-contained binary with no server to stand up, so the natural first production target is the scheduled work — the nightly refreshes and recurring reports that consume seats without needing a human at a terminal. That is where the per-seat saving lands soonest. Because both systems read and write the same file formats, you can move at whatever pace suits you.
| Feature | Jenner | SAS |
|---|---|---|
| Runs most SAS programs | Yes | Yes |
| Per-seat license required | No | Yes |
| Runs on Windows, macOS, Linux | Yes | Limited |
| Native CSV and Parquet I/O | Yes | No |
| Read SAS7BDAT and XPT | Yes | Yes |
| Write SAS7BDAT and XPT | Yes | Yes |
| Runs as a single self-contained binary | Yes | No |
| Native Jupyter kernel | Yes | No |
| Passes full NIST StRD benchmark (58/58) | Yes | Strong, with caveats |
Jenner implements 100 procedures using the same statement and option names as SAS, so a PROC call you already know works the way you expect it to. If your work depends on a specific procedure, the fastest way to judge the fit is to read its reference page — every procedure below links to its own documentation, covering the statements, options, and output it supports.
Coverage spans the core language, base and statistical procedures, PROC SQL, ODS Graphics, file formats, and database connectivity — 329 individual SAS behaviors documented, each procedure with its own reference page listing the statements and options it supports.
Start free — no SAS license required.