Released: 02.10.2026

New & Noteworthy

  • Seamless DL training workflow: better guidance and the ability to annotate in arivis Pro but train in arivis Cloud
  • Microscopy Copilot: your personal AI assistant, helps you learn all you ever wanted to know about arivis Pro
  • Use externally trained 3D semantic models (.onnx) in DL Segmenter for true volumetric segmentation
  • New Select tool providing multiple modes to select objects in the 2D and 4D Viewer

Analysis & AI

  • new: a new menu entry "AI Training" as a centralized way to our different AI trainers
  • new: AI Training screen provides detailed guidance to pick the right AI method
  • new: AI Home-screen - easily switch between our different AI trainer panels
  • new: DL Trainer allows to annotate in arivis Pro but train in arivis Cloud (for semantic or instance segmentation models)
  • new: contact handling options for drawn segments in the DL Trainer Draw tool - merge, overlap, touch
  • new: ability to upload annotated image data to arivis Cloud - full or partial (annotated areas only)
  • new: use single sign on (with ZEISS ID) for accessing your arivis Cloud account
  • new: DL Segmenter is able to run 3D ONNX models
  • change: Cellpose operation does not bring (partially outdated) models anymore, please download or train models on your own
  • new: Cellpose operation also supports Cellpose 4.x models (conversion to ONNX is required)
  • new: ability to define a default folder for your Cellpose modes for easier access in the operation
  • change: ML Trainer (Image) now uses scikit-learn as a backend
  • new: ML Trainer (Image) offers different models to choose from: Random Forest, Histogram Boosting, MLPC (Multi-layer perception classifier), PCA
  • new: tools of analysis operations (e.g. intensity or distance picker) work now in the 4D Viewer as well
  • improved: Distance operation calculations are much faster now
  • improved: resource utilization for Cellpose operation
  • improved: Data Aggregation window remembers the Hide empty wells setting
  • change: the DL Trainer environment has been updated to Python version 3.12.0
  • fix: tag auto complete works more reliable in Import Document Objects
  • fix: Distance operation does not write NaN into CSV exports anymore
  • fix: order of values in Oriented Bounds 3D
  • fix: proper sorting of wells in aggregation dialog and export
  • fix: progress for Object mask operation

Objects

  • new: aligned Select tool to select objects in the 2D and 4D Viewer
  • new: selection modes allow to choose from rectangle, brush or freehand (polygone)
  • new: brush selection mode allows to select objects by touching them with a resizable brush
  • new: freehand selection mode allows to select objects by drawing a polygone around them
  • improved: merge and fill holes behavior of the Draw Objects tool is not global anymore
  • improved: interpolation across planes now also works for the ellipse or polygon mode of the Draw Objects tool
  • improved: selection in 2D Viewer properly considers the actual outline of any segment, trace or track
  • improved: prevent accidentally deselection of objects on right click
  • improved: view "selected objects only" behaves the same for 2D and 4D Viewer now
  • change: bounding boxes for (selected) objects are turned off by default

4D Viewer

  • new: show handles for the clipping planes on mouse over in all tools of the 4D Viewer
  • new: allow to toggle visibility of the clipping plane handles via menu
  • improved: the Surface Quality parameter has been reworked and is now called Smoothing
  • improved: proper selection outline for multiple touching objects
  • improved: preserve trace shape and connectivity when converting traces to segments
  • improved: rotate by left click and drag in any of the 4D Viewer's tools while pressing the Alt key
  • improved: visibility of the ROI handles of the preview of an analysis operation
  • improved: object loading and meshing progress

Import & Data Handling

  • new: media reader to import JDCE data from Molecular Devices IN Carta 2
  • new: support for Yokogawa CQ3000 data import
  • new: ability to modify the pixel size for CZI data
  • new: ability to modify/write pixel data in CZI data
  • improved: better support for importing SmartSEM TIFF
  • improved: faster import of IMS data
  • improved: multi file Dicom (DCM) import (without DICOMDIR)
  • improved: better support for well plate metadata in CZI
  • improved: import of multi-page TIFF if they have different pixel types
  • fix: make sure to import all images from DICOMDIR
  • fix: IMS import for Tile Sorter now correctly combine channels
  • fix: CZI export writes proper time stamps for time series data
  • fix: some 16bit CZI data did not update properly on changes of the color range
  • fix: prevent a crash on importing Hamamatsu TIFF

General

  • new: Microscopy Copilot AI chatbot supports natural language questions for efficient use of arivis Pro and image analysis
  • improved: Well Picker panel modernized
  • improved: UI of the color bar with easier to use handles
  • change: arivis Pro runs on .NET 10 now
  • change: the integrated scripting environment has been updated to Python version 3.13.12
  • removed: menu action Copy Current Selection Content

Misc

  • improved: arivis Pro installed via ZMI will create a desktop icon
  • change: arivis Pro stand alone installer provides Docker Desktop version 4.89.0
  • various additional smaller bug fixes and general improvements

Deprecated Scripting API

Be prepared that these python API functions and classes will be removed in a future release:

  • arivis_analysis
    ** open_analysisui
    ** close_analysisui
    ** get_operation_registry
    ** Pipeline.delete_operation
  • arivis_core
    ** class Buffer
    ** class PointF2D
    ** class Point3DF
    ** class Size2D
    ** Document.add_imageset
    ** Document.add_pipeline
    ** ImageSet.get_bounding_rect
    ** ImageSet.get_pixel_width
    ** ImageSet.get_pixel_height
    ** ImageSet.get_pixel_depth
    ** ImageSet.get_channeldata
    ** ImageSet.set_channeldata
    ** ImageSet.set_gradient
    ** ImageSet.get_gradient
    ** ImageSet.set_gamma
    ** ImageSet.get_gamma
    ** ImageSet.set_color_range
    ** ImageSet.get_color_range
  • arivis_objects
    ** Marker.set_simple_location
    ** Marker.set_x
    ** Marker.set_y
    ** Marker.set_z
    ** Marker.set_locationf
    ** Marker.set_simple_sublocation