v1.0.0 v6.14.1

high risk

imbhargav5/open-recorderSep 6, 2026 → Sep 6, 20260 days between releases

What's New

Newer release adds workflow screenshots and gallery middle-click features, while older release focused on unified realtime commands.

Breaking Changes

  • The older release's commands (`serve`, `talk`, `local`) are not present in the newer release.
  • The older release's speech backends (Qwen3, OpenAI, vLLM) are not present in the newer release.
  • The older release's `--log_transcripts` flag is not present in the newer release.
  • The older release's Apple Silicon setup is not present in the newer release.
  • The older release's OpenAI Agents SDK integration is not present in the newer release.

New Features

  • Workflow screenshot saving
  • Middle-click gallery feature
  • Krea-2 DoRA support

Improvements

  • VRAM memory reservation improvements for VAEs
  • Internal memory management at shutdown time

Bug Fixes

  • Robust handling of Kohya keys
  • Security hole fix for AlibabaCloud image downloads
  • Truncated image moves fix

Upgrade Notes

The newer release is a different project (InvokeAI) and does not contain the commands or backends from the older release (speech-to-speech).

The Two Releases

v1.0.0

v1.0.0

Sep 6, 2026· 2 months ago

The first major release introduces a unified Realtime engine with explicit commands (`serve`, `talk`, `local`), expanded speech backends, and more reliable conversation lifecycle handling.

Full release notes
v6.14.1

Version 6.14.1

Sep 6, 2026· 2 months ago

Fixes bugs and adds workflow screenshot saving and middle-click gallery features to InvokeAI.

Full release notes