WEBKNOSSOS
Sign up

VISUALIZE · ANNOTATE · SHARE · ANALYZE WITH AI

VISUALIZE · SHARE · ANNOTATE · ANALYzE

Large-scale 3D microscopy, right in your browser

Large-scale 3D microscopy, right in your browser

WEBKNOSSOS lets labs explore, annotate, and reconstruct volume electron and light microscopy datasets, from a single volume to petabyte-scale connectomics. Share a dataset with a link, and divide the work across your team.

VISUALIZE

See your data at full resolution, in the browser.

Open a dataset and move through it at native resolution, from gigabytes to petabytes, with nothing to install. Built for volume EM, and at home with light microscopy, X-ray, and correlative microscopy.

Learn more about data sources, working with meshes, and correlating datasets.

Data: Berg et al., bioRxiv 2025.

VISUALIZE

See your data at full resolution, in the browser.

Open a dataset and move through it at native resolution, from gigabytes to petabytes, with nothing to install. Built for volume EM, and at home with light microscopy, X-ray, and correlative microscopy.

Learn more about data sources, working with meshes, and correlating datasets.
ANNOTATE

Trace and segment in one workspace.

Skeleton tracing and volume segmentation side by side, with AI quick-select and interpolation speeding up the slow parts. Every change autosaves, with full version history.

Learn more about skeleton & volume annotations, and segment measurements.

Data: Motta et al., Science 2019.

ANNOTATE

Trace and segment in one workspace.

Skeleton tracing and volume segmentation side by side, with AI quick-select and interpolation speeding up the slow parts. Every change autosaves, with full version history.

Learn more about skeleton & volume annotations, and segment measurements.
COLLABORATE

Share the data, split the work.

Send a link that opens the exact view you’re looking at. No downloads, no file versions. Work on the same datasets as your colleagues, and divide large annotation jobs across the team.

Learn more about users and permissions, and collaboration in WEBKNOSSOS.

Data: Motta et al., Science 2019.

COLLABORATE

Share the data, split the work.

Send a link that opens the exact view you’re looking at. No downloads, no file versions. Work on the same datasets as your colleagues, and divide large annotation jobs across the team.

Learn more about users and permissions, and collaboration in WEBKNOSSOS.
ANALYZE WITH AI

Segment automatically, proofread together

Run pre-trained models to segment neurons and mitochondria, or train your own. Then proofread the result together, turning a rough segmentation into a connectome you can publish.

ANALYZE WITH AI

Segment automatically, proofread together

Run pre-trained models to segment neurons and mitochondria, or train your own. Then proofread the result together, turning a rough segmentation into a connectome you can publish.

WORKFLOW

Supporting your work from raw images to analysis insights

Make WEBKNOSSOS the central hub for large image data in your lab. Import data from your microscopes. Visualize, manage and collaboratively annotate your data directly in WEBKNOSSOS. Stream your data to other tools for advanced analysis via OME-Zarr support.

WORKFLOW

Supporting your work from raw images to analysis insights

Make WEBKNOSSOS the central hub for large image data in your lab. Import data from your microscopes. Visualize, manage and collaboratively annotate your data directly in WEBKNOSSOS. Stream your data to other tools for advanced analysis via OME-Zarr support.

Acquire
Import from SEM, multi-SEM, and light microscopy, then combine modalities in one place.
Align
Align and register serial sections directly in WEBKNOSSOS, before you segment. No separate tool.
Annotate
Paint volumes with AI quick-select and trace skeletons in the same workspace. Work on one annotation together, each from your own computer.
Analyze
Segment neurons and mitochondria with pre-trained models, or train your own on your annotations. No pipeline to build.
Proofread
Fix split and merge errors on large segmentations by editing the supervoxel graph, collaboratively and in near real time.
Publish
Share a dataset with a link, with access control so it stays private until you choose. Export publication-ready animations and figures.
Interoperability
Stream your data out as OME-Zarr so tools like napari, Fiji, and Neuroglancer can read it. Script and automate everything through the Python library and REST API.
Zarr
Load OME-Zarr (NGFF) datasets, and export your data back out as a Zarr stream. WEBKNOSSOS also reads WKW, N5, and Neuroglancer Precomputed.
Cloud storage
Stream datasets directly from Amazon S3, Google Cloud Storage, or an HTTP server. Your data stays where it is and loads read-only, without counting against your storage quota.
WEBKNOSSOS workflow

CASE STUDIES

Published research runs on WEBKNOSSOS

Labs use WEBKNOSSOS to get from raw data to results that reach the literature. Here are two: a large-scale connectomic comparison, and a correlative microscopy study across seven modalities.

CASE STUDIES

Published research runs on WEBKNOSSOS

Labs use WEBKNOSSOS to get from raw data to results that reach the literature. Here are two: a large-scale connectomic comparison, and a correlative microscopy study across seven modalities.

CONNECTOMICS
Routine connectomic reconstructions
Loomba et al. reconstructed nine datasets fully automatically in a few days, revealing differences in the interneuron network between mouse and human.
Learn more
CONNECTOMICS
Routine connectomic reconstructions
Loomba et al. reconstructed nine datasets fully automatically in a few days, revealing differences in the interneuron network between mouse and human.
Learn more
CORRELATIVE MICROSCOPY
Sample prep and warping accuracy in the olfactory bulb
After imaging one brain with seven modalities, Zhang et al. explored how to generate reliable correlative maps, and the central role WEBKNOSSOS played in their workflow.
Learn more
CORRELATIVE MICROSCOPY
Sample prep and warping accuracy in the olfactory bulb
After imaging one brain with seven modalities, Zhang et al. explored how to generate reliable correlative maps, and the central role WEBKNOSSOS played in their workflow.
Learn more

GLOBAL CONNECTOMICS COMMUNITY

Trusted by labs worldwide

From university cores to national facilities, WEBKNOSSOS keeps massive microscopy datasets accessible where collaboration happens.

GLOBAL CONNECTOMICS COMMUNITY

Trusted by labs worldwide

From university cores to national facilities, WEBKNOSSOS keeps massive microscopy datasets accessible where collaboration happens.

“WEBKNOSSOS is the best web-based tool out there for integrated team-based annotation, proofreading, display, and output of serial EM data.” –
David Berson
Professor at Brown University
“WEBKNOSSOS enables extraordinary ease in team annotations of our connectomic datasets.” –
Gregg Wildenberg
Staff Scientist, University of Chicago
“WEBKNOSSOS is very useful for safely storing large datasets and managing access across multiple users.” –
Carles Bosch
Principal Laboratory Research Scientist, Francis Crick Institute
“With WEBKNOSSOS, I can visualize the ground truth and my predictions (skeletons or volume annotations) as overlapping layers.” – Read the article.
Valentin Pinkau
Data scientist at scalable minds
“We use WEBKNOSSOS a lot for data visualization, annotation, and result evaluation. It is our primary tool for volume annotations.” – Read the article.
Sahil Loomba
Post-doc research scientist, Max Planck Institute for Brain Research
“WEBKNOSSOS played a central role in our research. We stored all of our datasets there and used it for data sharing, collaboration, and annotation.”
Luke Noon
Team leader of the Metabolic Growth Signals and Regenerative Medicine Laboratory
“WEBKNOSSOS is the best web-based tool out there for integrated team-based annotation, proofreading, display, and output of serial EM data.” –
David Berson
Professor at Brown University
“WEBKNOSSOS enables extraordinary ease in team annotations of our connectomic datasets.” –
Gregg Wildenberg
Staff Scientist, University of Chicago
“WEBKNOSSOS is very useful for safely storing large datasets and managing access across multiple users.” –
Carles Bosch
Principal Laboratory Research Scientist, Francis Crick Institute
“With WEBKNOSSOS, I can visualize the ground truth and my predictions (skeletons or volume annotations) as overlapping layers.” – Read the article.
Valentin Pinkau
Data scientist at scalable minds
“We use WEBKNOSSOS a lot for data visualization, annotation, and result evaluation. It is our primary tool for volume annotations.” – Read the article.
Sahil Loomba
Post-doc research scientist, Max Planck Institute for Brain Research
“WEBKNOSSOS played a central role in our research. We stored all of our datasets there and used it for data sharing, collaboration, and annotation.”
Luke Noon
Team leader of the Metabolic Growth Signals and Regenerative Medicine Laboratory

RESEARCH & PRODUCT

Highlights from our blog

Tutorials, feature deep-dives, and stories from the labs using WEBKNOSSOS.

RESEARCH & PRODUCT

Highlights from our blog

Tutorials, feature deep-dives, and stories from the labs using WEBKNOSSOS.

TUTORIAL
Making a flythrough video of your data with WEBKNOSSOS and your AI coding agent
September 2026
Turn your data into beautiful, professional videos. Much easier now with AI coding agents like Claude.
FEATURE
Maximum Intensity Projection now in WEBKNOSSOS
August 2026
MIP renders your dataset directly in the 3D viewport: a spatial view like meshes, but for any image layer, not just segmentations.
INTERVIEW
Teaching Cell Biology with Real EM Data: An Interview with Ahmed Elewa, Miami University
May 2026
How researchers and students explore large 3D microscopy datasets directly in the browser.
CASE STUDY
From Beamline to Browser: Automating High-Throughput X-ray Imaging at EMBL P14
March 2026
A fully automated pipeline transfers reconstructed synchrotron X-ray tomography datasets straight into a browser-based 3D environment.
TUTORIAL
Custom model training for mitochondria detection in WEBKNOSSOS
February 2026
We trained a custom instance segmentation model on a real volume EM dataset and deployed it across the full 140 GB volume.
TEACHING
Cell biology exercises with WEBKNOSSOS
October 2024
Students explore the brain in 3D at high resolution, identify cell processes, and learn to recognize synapses.
TEACHING
Neuroanatomy exercises with WEBKNOSSOS
September 2024
Exercises designed by Prof. Berson (Brown University) introduce students to connectomic analysis of retinal neurons.
FEATURE
AI Quick-Select Tool
September 2024
Quick-select now works in 3D and predicts up to 16 sections. A single click segments a cell, no rectangle needed.
TUTORIAL
Making a flythrough video of your data with WEBKNOSSOS and your AI coding agent
September 2026
Turn your data into beautiful, professional videos. Much easier now with AI coding agents like Claude.
FEATURE
Maximum Intensity Projection now in WEBKNOSSOS
August 2026
MIP renders your dataset directly in the 3D viewport: a spatial view like meshes, but for any image layer, not just segmentations.
INTERVIEW
Teaching Cell Biology with Real EM Data: An Interview with Ahmed Elewa, Miami University
May 2026
How researchers and students explore large 3D microscopy datasets directly in the browser.
CASE STUDY
From Beamline to Browser: Automating High-Throughput X-ray Imaging at EMBL P14
March 2026
A fully automated pipeline transfers reconstructed synchrotron X-ray tomography datasets straight into a browser-based 3D environment.
TUTORIAL
Custom model training for mitochondria detection in WEBKNOSSOS
February 2026
We trained a custom instance segmentation model on a real volume EM dataset and deployed it across the full 140 GB volume.
TEACHING
Cell biology exercises with WEBKNOSSOS
October 2024
Students explore the brain in 3D at high resolution, identify cell processes, and learn to recognize synapses.
TEACHING
Neuroanatomy exercises with WEBKNOSSOS
September 2024
Exercises designed by Prof. Berson (Brown University) introduce students to connectomic analysis of retinal neurons.
FEATURE
AI Quick-Select Tool
September 2024
Quick-select now works in 3D and predicts up to 16 sections. A single click segments a cell, no rectangle needed.
WEBKNOSSOS

Annotate, proofread, and analyze large 3D microscopy datasets in the browser.

WEBKNOSSOS

Annotate, proofread, and analyze large 3D microscopy datasets in the browser.