idtracker.ai tracks up to 100 unmarked animals from videos recorded in laboratory conditions using artificial intelligence. Free and open source.

This work has been published in eLife Methods, please include the following reference if you use this software in your research:

• plain Jordi Torrents, Tiago Costa, Gonzalo G de Polavieja. New idtracker.ai: rethinking multi-animal tracking as a representation learning problem to increase accuracy and reduce tracking times. Life14:RP107602 (2026)
• bibtex @article{idtrackerai_2026, title={New idtracker.ai: rethinking multi-animal tracking as a representation learning problem to increase accuracy and reduce tracking times}, url={https://doi.org/10.7554/eLife.107602}, DOI={10.7554/elife.107602}, publisher={eLife Sciences Publications, Ltd}, journal={eLife}, author={Torrents, Jordi and Costa, Tiago and de Polavieja, Gonzalo G}, year={2026} }

Visit our website to find more information about the software, installation instructions, and user guides.

Installation for developers

On an environment with Python>=3.10 and a working installation of Pytorch (Torch and Torchvision) you can install the latest published idtracker.ai version by installing directly form the GitLab repo:

pip install git+https://gitlab.com/polavieja_lab/idtrackerai

Or install the developing version from the develop branch:

pip install git+https://gitlab.com/polavieja_lab/idtrackerai@develop

There exist two extra dependencies options:

• dev to install tools for formatting, static analysis, building, publishing, etc.
• docs to install needed packages to build documentation (sphinx and some plugins).

Contributors

• Jordi Torrents (2022-)
• Tiago Costa (2024)
• Antonio Ortega (2021-2023)
• Francisco Romero-Ferrero (2015-2022)
• Mattia G. Bergomi (2015-2018)
• Ricardo Ribeiro (2018-2020)
• Francisco J.H. Heras (2015-2022)

For more information please send an email (info@idtracker.ai) or use the tools available at https://gitlab.com/polavieja_lab/idtrackerai.