v0.24.1

Release date: 2026-06-24

Early Access

The Luminary Cloud API and Python SDK are Early Access features that are still under development. View the Luminary Cloud Early Access Terms. The interface may change, and you may need to migrate or discard scripts and configuration files multiple times before v1.0.0 is released. In the future, migrations may also be necessary when upgrading to new major versions of the SDK (e.g. v2.0.0).

Release Notes

This release adds pipeline output types for rendered images and data extracts, and per-pixel buffer outputs to the image renderer.

Breaking Changes

None.

Features

  • New pipeline output types PipelineOutputImages and PipelineOutputVisDataExtract allow pipeline script stages to produce rendered images or data extracts as stage outputs. Image outputs are displayed in the pipeline UI.

  • New vis.get_render_output() convenience function retrieves a RenderOutput by its extract ID alone (no project ID required).

  • New vis.get_data_extract() convenience function retrieves a DataExtract by its extract ID alone (no project ID required).

  • New Scene.render_outputs() method requests a render that produces per-pixel numeric buffers in addition to the RGB image. Pass scalar=True to get the scalar field buffer and/or world_position=True to get the 3-D world-space position buffer. The object-id buffer is always included alongside the requested buffers.

  • New RenderOutput.download_outputs() method downloads the results of a buffer render, returning one RenderResult per camera. Each RenderResult carries the RGB image plus an OutputBuffers object containing the object-id buffer and any requested scalar/position buffers (see BufferData). The object-id buffer maps each pixel to a filter and primitive, enabling per-surface attribution. Background pixels carry INVALID_OBJECT_ID.

    Note: This feature is experimental and may change in a future release.

Bug Fixes / Improvements

None.

Migration Guide

No breaking changes.