Antscan opens a free 3D atlas of 800 ant species to researchers and the public

The shape of an organism is one of the oldest subjects in biology and one of the hardest to study at scale. For small creatures like ants -- some barely visible to the naked eye -- capturing accurate three-dimensional morphology has required expensive, time-consuming micro-CT scans that produce exquisite data one specimen at a time. Scaling that process to hundreds of species has been, until now, impractical.

Antscan changes the math. A project co-led by researchers at the Okinawa Institute of Science and Technology (OIST) in Japan and the Karlsruhe Institute of Technology (KIT) in Germany has produced an open-access database of more than 2,000 three-dimensional ant models representing 800 different species. A paper in Nature Methods describes both the data and the workflow used to create it.

Speed through particle physics

The bottleneck in conventional micro-CT scanning is time. A single specimen can take 10 hours to image in a standard lab-based scanner. At that rate, scanning 2,000 ants would require approximately six years of continuous operation.

The Antscan team compressed that timeline by using a synchrotron particle accelerator at KIT. The synchrotron produces an X-ray beam far more intense than any lab source, allowing each specimen to be scanned in about 30 seconds. A robotic sample changer rotated and swapped specimen vials automatically, producing 3,000 two-dimensional X-ray images per ant that were then computationally reconstructed into three-dimensional models.

"If we were to carry out this project with a lab-based CT scanner, it would take around six years of continuous operation. With our setup, we scanned 2,000 specimens in a single week," said first author Dr. Julian Katzke, a graduate of OIST.

Beyond the exoskeleton

The scans go far deeper than surface appearance. At micrometer resolution, they reveal internal anatomy: muscles, nervous systems, digestive tracts, and stingers are all visible. This level of detail enables research into biomechanics, comparative anatomy, and evolutionary morphology that would otherwise require destructive dissection of rare specimens.

The models can also be animated. With the exoskeleton and musculature captured in high resolution, scientists and artists can model ant movement for research or realistic multimedia depictions.

Curation before computation

The high-tech scanning was preceded by decidedly low-tech labor. OIST coordinated the collection of ethanol-preserved ant specimens from institutions, museum collections, and expert collectors around the world. Each specimen was sorted by species and caste, and its metadata -- who collected it, where, when -- was standardized to ensure accurate labeling.

"This work moves us further into the big data era of capturing, analyzing, and sharing organismal shape and form," said Prof. Evan Economo of OIST's Biodiversity and Biocomplexity Unit. "The potential for integrating this data is immense and very exciting."

Already generating results

The database has already been used in published research. A December 2025 study in Science Advances drew on Antscan data to investigate the relationship between exoskeleton thickness and colony size across more than 500 ant species. The finding: ant colonies that invest less in individual armor produce larger, more diverse societies. Exoskeleton thickness -- measured via cuticle volume from the 3D scans -- had been extremely difficult to quantify before Antscan made it possible at scale.

Combined with high-quality ant genomes published in Cell in 2025, the morphological data enables new kinds of research connecting physical form to genetic variation across the ant evolutionary tree.

Free and open to everyone

All raw scan files are freely downloadable, and the Antscan portal includes a built-in 3D viewer for each specimen. The openness is intentional. High-resolution micro-CT scans are expensive to produce, and smaller institutions, citizen scientists, and educators have historically been priced out of access.

"To me, that's the most exciting part of the project: opening up the database to a potentially infinite variety of perspectives," Katzke said. "I'm thrilled to see how other people will use this data in ways that I couldn't have imagined."

The project also serves as a template. The same synchrotron-robotics-AI workflow could be applied to any group of small organisms preserved in the world's natural history collections: beetles, flies, spiders, wasps. The infrastructure exists; what is needed is the curatorial work of selecting, sorting, and shipping specimens.

For a field that has long struggled to scale morphological data collection, Antscan demonstrates that the technical barriers are no longer the limiting factor.