Cellpose
Generalist deep-learning cell and nucleus segmentation
What's new
Cellpose 4 (2026) adds DINOv3-based models (cpdino) and an updated Cellpose-SAM (cpsam_v2) that produces fewer spurious masks in low-contrast regions.
Cellpose is a generalist segmentation model for cells and nuclei that works across imaging modalities without retraining — though it can be fine-tuned on your own data.
Highlights
- Robust to shot noise, blur, undersampling and contrast inversions.
- Works in 2D and 3D, with a napari-friendly GUI and a Python API.
- Cellpose 4 introduces Cellpose-SAM and DINOv3 backbones.
Example
1from cellpose import models2model = models.CellposeModel(gpu=True)3masks, flows, styles = model.eval(img, diameter=None)
Related tools
Open source
Bioimage Analysis
napari
Fast, interactive n-dimensional image viewer for Python
Python libraryDesktop app
Open source
Bioimage Analysis
StarDist
Star-convex object detection for nuclei segmentation
Python libraryDesktop app
Open source
Bioimage Analysis
Fiji / ImageJ
"Fiji Is Just ImageJ" — the classic bioimage analysis platform
Desktop app