Natively

Natively combines desktop audio transcription, reference-file retrieval and model-generated answers for meetings and interviews. The English documentation offers macOS and Windows downloads, plus a macOS Homebrew route. It supports local speech models and describes an offline configuration using Ollama; cloud inference depends on the user’s configured provider and credentials.

Its current Personal Use Source License permits personal, educational, research and non-commercial use. It is source-available rather than permissively licensed open-source software. Commercial use requires separate written permission. The October 15 source change addresses local model shutdown and interrupted downloads; that commit should not be mistaken for a newly published stable installer or an independently verified reliability result.