All features
Labrix feature guide

Experiment tracking

Log runs, parameters, seeds, environment and outcomes, then reuse those records when you write the Methods section.

Start free
A researcher reviews a scientific paper beside a laptop in a university lab.

What it adds to the project

Concrete capabilities, kept close to the sources and decisions they support.

Experiments and their runs, with parameters, seeds and environment captured per run

Code references linking a run to the commit that produced it

Results recorded next to the settings that generated them, not in a separate spreadsheet

Methods and settings flow into your manuscript instead of being reconstructed from memory

Shared with your lab, so a run is still legible after the person who ran it graduates

How it works

A short path from your material to something you can inspect, revise and reuse.

Create the experiment

One experiment per research question. It becomes the container for every run you try against it.

Log each run

Record parameters, seed, environment, the code reference and the outcome. A failed run is worth logging - it is what stops you repeating it.

Write from the record

When you draft the paper, the Methods and experimental-setup details come from what you logged, so the reported protocol matches the one you actually ran.

Common questions

Is this a replacement for Weights & Biases or MLflow?
Different job. Those instrument training loops and stream metrics. Labrix tracks experiments at the level a paper needs - settings, seeds, environment, outcome - and connects them to the manuscript. Many labs run both.
Does it work for wet-lab work?
Yes. Runs are not assumed to be code: a protocol, a reagent batch, a cohort or an instrument setting fits the same structure.
Can my lab see my runs?
If you share them. Team members can view and edit shared records; deleting stays with the owner.

Try it with your own material

Labrix is free to start, with no credit card required for the basics.

Start free