Building the Future of Atmospheric Retrievals

Advancing exoplanet science with TauREx3, a state-of-the-art retrieval framework

Our Vision

We envision a future where the atmospheres of distant worlds can be understood with clarity and precision. TauREx3 — a fully Bayesian, modular atmospheric retrieval code — is at the heart of the exoplanet revolution. Available for all, TauREx is flexibility and easy to use: it allows scientists to rapidly simulate and analyze spectra, unlocking insights into temperature, pressure, and chemical compositions of exoplanet atmospheres.

While mostly using intuitive code in python, TauREx does not sacrifice performances! The code is highly optimized and HPC ready: cross-node ready up to 1000+ CPUs without performance loss, and GPU-accelerates. Version 3.0 offers tenfold increase in forward-modeling speed and retrieval times reduced by a factor of six compared to previous versions. TauREx modular architecture lets users flexibly define atmospheric parameters, incorporate custom physics, and explore planetary scenarios with unprecedented efficiency.

By combining robust Bayesian inference with transparent, open-source design, TauREx 3 empowers the exoplanet community to push the boundaries of atmospheric science — from current observations to the next generation of space telescopes.

The TauREx Team

Ahmed F. Al-Refaie

Core Team Member

Quentin Changeat

Core Team Member

Sushuang Ma

Core Team Member

Gordon Yip

Core Team Member

Used by Researchers at

TauREx Core Values

Community

We intend TauREx for everyone. That's why TauREx is coded in python and is the first retrieval to offers a plugin system for easier community inputs.

Performance

We use TauREx for state-of-the art research. That's why we need the best! TauREx has been designed with performance at its core and will work best on large HPC architectures.

Transparency

We believe in open science and building reproducible research through transparency. That's why TauREx is delivered via Open-Source BSD3 License