Stella brings local semantic search to macOS
Developed by Waterloo engineers, Stella is a local, privacy-focused semantic search application for macOS that indexes local documents to allow natural language queries instead of exact filename matches. The tool runs entirely on-device, packaging all necessary models into a 1.5 GB installer to guarantee offline functionality and zero cloud dependence.
Stella's offline, local-first architecture is a compelling alternative to cloud-reliant indexing tools, presenting a secure, zero-trust solution for users handling sensitive files.
- –**True Semantic Indexing**: By searching document concepts and meanings rather than simple metadata or filenames, Stella effectively solves a major daily workflow bottleneck that traditional search tools like Spotlight fail to address.
- –**Privacy and Security**: Packaging the AI models locally ensures no data leaves the user's machine, satisfying strict privacy needs of corporate, legal, and personal workflows.
- –**Native User Experience**: Features like custom system-wide keyboard shortcuts and direct drag-and-drop into target applications like Gmail make it highly practical and integrated.
- –**System Constraints**: The current beta is exclusive to macOS Apple Silicon (M1/M2/M3) and requires a hefty 1.5 GB initial download, temporarily alienating Windows and Intel Mac users.
DISCOVERED
1h ago
2026-06-01
PUBLISHED
6h ago
2026-06-01
RELEVANCE
AUTHOR
[REDACTED]