Changelog#

0.3.10 (2026-02-19)#

New:

  • Enable the use of output scalars in global coefficient formula.

  • Added support for generating, managing, and logging in with offline tokens via the current user object SimAIClient.me.

  • Add new set of methods to access lists of objects for Workspace, Project, Geometry and GeomAIProject

Changes:

0.3.9 (2026-01-21)#

Changes:

0.3.8 (2026-01-16)#

New:

  • Make the nb_latent_param field in GeomAIModelConfiguration optional, and update its default and maximum value.

  • Added rename method for training data.

Fixes:

  • Replace truthy fallback with explicit None check.

0.3.7 (2025-12-17)#

New:

  • Add basic SimAI examples in documentation.

  • Add method GeomAIWorkspace.get_latent_parameters() to retrieve latent parameters as dict in GeomAIWorkspace.

  • Remove deprecated Optimization.run_parametric() optimization method.

  • Add model_configuration property in Workspace and GeomAIWorkspace.

  • Allow users to use scalars in model output in model configuration and add a new post-processing PredictScalars.

Changes:

  • Increase default timeout for long operations.

  • Deprecate boundary_conditions in favor of scalars.

0.3.6 (2025-12-03)#

New:

Changes:

Fixes:

  • Fix post-processing inputs in model configuration not being included in global coefficients processing.

  • Fix incorrect handling of dictionaries when building a GeomAI model.

0.3.5 (2025-10-14)#

New:

  • Add GeomAI model evaluation report download capabilities

  • Add support for point data in calculette and model training

Changes:

  • Replace requests/niquests with httpx

  • Documentation improvements

  • Refactor the authentication logic

Fixes:

  • Fix GeomAI prediction deletion

0.3.4 (2025-09-24)#

New:

  • Add axial_symmetry parameter to non-parametric optimization

Changes:

  • Non-parametric optimization now returns OptimizationResult

  • Switch to Niquest built-in SSE support

  • Re-order non-parametric optimization parameters

  • Improve error message when creating client with wrong organization

  • Documentation improvements

Fixes:

  • Fix improper exception handling

  • Fix unhandled SSE error

0.3.3 (2025-07-02)#

Fixes:

  • Replace verify_gc_formula with process_gc_formula.

0.3.2 (2025-07-02)#

Changes:

  • General GeomAI improvements, including docs

0.3.1 (2025-06-24)#

New:

  • Add max_displacement parameter to non-parametric optimization

Changes:

  • Add checks on build on top

  • Boundary conditions are optional on optimization runs

  • Use new way to process global coefficients formula that uses a cache

  • Improved checks for parametric and non parametric optimizations

  • Switch to uv

  • Docs improvements

0.3.0 (2025-05-14)#

New:

  • Retry sending requests on HTTP 5xx error (#135)

Changes:

  • Include python version in user_agent (#143)

  • SSE: Don’t expect record from jobs with pending status (#141)

0.2.7 (2025-03-26)#

New:

Fixes:

  • Stop using deprecated endpoints

  • Deleting PointCloud now cleanses the CustomVolumePointCloud post-processing cache

0.2.6 (2025-01-23)#

New:

Changes:

Fixes:

  • Type hints on SimAIClient off by one

  • Fix pysimai version check

0.2.5 (2024-11-05)#

New:

Changes:

  • Optimization.run() is now Optimization.run_parametric() and checks that the generation function has a suitable signature

  • Remove deprecated design of experiments feature

  • Resolution steps are now printed upon error if any

Fixes:

  • Correct payload for surface post-processing inputs on model build

0.2.4 (2024-09-23)#

New:

  • Auth tokens are now cached in file system and get re-authenticated in a parallel fashion.

  • Invalid refresh token now triggers a reauth instead of crashing.

  • build_preset option in ModelConfiguration can now be one of debug, 1_day, 2_days, 7_days.

  • Model Evaluation Report data (csv file) can now be downloaded with download_mer_data.

  • Typing improvements; introducing JSON type is introduced and APIResponse type is updated to include JSON type.

  • New property Prediction.raw_confidence_score is added to Prediction, which returns the raw confidence score.

Fix:

  • Fixed the error where data was not in coordinance with the BE response. data now runs without errors.

0.2.3 (2024-08-21)#

New:

  • Added PostProcessInput class to define post processing input fields.

  • Added support for NaN and Inf for Global Coefficients and Post Processings.

Fixes:

  • Removed compute argument from TrainingData.upload_folder()

  • Fixed Model Configuration to raise a ProcessingError when volume field is missing from a sample specifying volume output.

  • Removed wakepy error mode success (deprecated) during optimization.

  • Renamed TrainingData method compute() to TrainingData.extract_data().

  • Updated documentation of GeometryDirectory.upload(): the workspace_id argument was moved to workspace but never updated.

0.2.2 (2024-07-17)#

New:

Fix:

  • Remove internal uses of deprecated workspace.model

0.2.1 (2024-06-28)#

Fixes:

0.2.0 (2024-06-28)#

New:

  • Model configuration can now be created from scratch and be used in training requests.

  • Training-data subsets can now be assigned to None. Options Ignored and Validation are retired.

Fixes:

  • Fixed bug when uploading large files. Large files can now be uploaded.

  • Fixed bug when listing prediction without current_workspace being set.

0.1.7 (2024-04-30)#

New:

  • Added DomainOfAnalysis class to help set the domain of analysis on a new model.

  • Add workspace option where we previously relied only on the global workspace

  • Add prediction.post.list()

Fixes:

  • Reestablish python 3.9 compatibility.

  • Bump wakepy lib to fix errors when not able to prevent sleep during optimization.

0.1.6 (2024-04-25)#

New:

Fixes:

  • The method Optimization.run()<ansys.simai.core.data.optimizations.OptimizationDirectory.run> now raises an exception if no workspace is provided and none is configured.

  • Fix RecursionError on authentication refresh

0.1.5 (2024-04-15)#

  • Training can now be launched using the most recent model configuration from a project.

  • Enabled non-interactive mode capability, allowing for automation or operations without manual inputs.

  • Added new validation Project.is_trainable() to verify if the project meets all minimum requirements for training.

  • Added new method Project.get_variables() to get all available variables used for a model’s inputs and outputs.

  • Fixed bug where a subset of training data could not be pulled. A subset of training data is now correctly retrieved.

  • Fixed erroneous call to a private function during the optimization run.

0.1.4 (2024-02-26)#

  • Less verbose sse disconnects

  • Fix client config vars being described two times

  • Fix type/KeyError in workspace.model.post_processings

  • Fix monitor_callback interface not respected in upload_file_with_presigned_post

  • Fix README indentation

0.1.3 (2024-02-02)#

Fix config args not taken into account if a config file is not found

0.1.2 (2024-01-24)#

Fix training data upload_folder method

0.1.1 (2024-01-19)#

Fix badges

0.1.0 (2024-01-19)#

Initial release