Record Genius Changelog

Stay updated with the latest features and improvements in Record Genius.

v1.0.2:

  • Updated the drop window to allow users to copy and paste file paths and browse from their computer path
  • Refactored the audio processing - now, audio will be converted into MIDI, sent for suggestions, and then data is returned and processed as a MIDI loop for the user.
  • Added a drop-down menu to choose the track type - this allows AI to be more specific in its recommendations and allows the audio process.
  • Updated the MIDI processing so the auto-generated suggestions save as intended.
  • Fixed an occasional freeze when exporting generated loops.

v1.0.3

  • Updated the MIDI processing to correctly parse note durations

v1.0.4

  • Added a 60-second timeout on all requests
  • Added Gemini, Claude, DeepSeek, and Grok models
  • Added a settings tab to allow users to change the model and learn about Record Genius
  • Updated error handling

v1.0.5

  • Added an upload-length limit (4-bars and 10 seconds)
  • Updated the settings tab
  • Updated the saving/load of settings
  • Removed claude from the AI model options (cost inefficient)
  • Changed the default AI model to grok
  • Updated authentication requests

v1.0.6

  • Added local LLM options: tinyllama, gemma3:1b, and llama3.2 (must have [ollama](https://ollama.com/) installed and running)
  • Updated AI responses to ensure correct MIDI processing
  • Updated the MIDI processing to track processed notes more efficiently

v1.0.7

  • Added Windows support
  • Updated API-Key authentication and the Settings tab with an API-Key field
  • updated the Chat tab with modes Mix, Ask, Brainstorm and prompt suggestions for each

This introduces a Record Genius app which is used to communicate to the plugin to control faders and more to autonomously mix your track on command

Support