Starstruk Logo MarkStarstruk

v0.61.0

roboflow/inferencev0.61.0Nov 17, 2025by grzegorz-roboflow

AI Summary

This release enables email sending via workflows using Roboflow's email relay and introduces WebRTC data decoupling to separate video and data channels for bandwidth efficiency. It adds a Dinov3 classifier and makes GPU optional for CPU-only functions. Maintenance updates include workflow validation and improvements to OPC-UA integration.

Key Highlights

  • API key passthrough/email for workflows
  • WebRTC data decoupling
  • Dinov3 classifier
  • GPU optional for CPU-only functions

New Features

  • Email Relay
  • WebRTC Data Decoupling
  • Dinov3 Classifier
  • OPC-UA Improvements

Full Release Notes

# 💪 Added
* Api key passthrough/email - with this change @yeldarby enabled our users to send emails from workflows using Roboflow's email relay - change by @yeldarby in https://github.com/roboflow/inference/pull/1685
* WebRTC data decoupling - video channel and data channel are now created independently, making it possible to process video without streaming preview back to preserve bandwidth - change by @lrosemberg in https://github.com/roboflow/inference/pull/1700
* Dinov3 classifier by @mkaic in https://github.com/roboflow/inference/pull/1705

# 🔧 Fixed
* Fix exp integration tests / re-enable clip baseline tests by @hansent in https://github.com/roboflow/inference/pull/1716
* Add missing 'workflows' marker to pytest.ini by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1724
* Handle Supervision 0.27.0 by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1725

# 🏡 Maintenance
* Validate workflow accepts api key by @SolomonLake in https://github.com/roboflow/inference/pull/1697
* switch modal to app cls by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1707
* Parametrize gpu type by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1709
* Parametrize webrtc worker timeout by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1708
* Make GPU optional and allow user to run on CPU-only functions by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1713
* OPC-UA improvements for ignition by @sberan in https://github.com/roboflow/inference/pull/1706
* OPC-UI write block: support additional type conversions by @sberan in https://github.com/roboflow/inference/pull/1719
* Accept newest numpy in inference dependencies by @PawelPeczek-Roboflow in https://github.com/roboflow/inference/pull/1721
* Add shared thread-pool to HTTP API workflows runner by @PawelPeczek-Roboflow in https://github.com/roboflow/inference/pull/1717
* OPC-UA Writer Block: Convert incoming values to primitive type prior to setting value by @sberan in https://github.com/roboflow/inference/pull/1723
* Modal parametrization / webrtc turn config / perspective correction fix by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1722

## New Contributors
* @mkaic made their first contribution in https://github.com/roboflow/inference/pull/1705

**Full Changelog**: https://github.com/roboflow/inference/compare/v0.60.0...v0.61.0