Show HN: Sageling – free, private, local, Cowork AI harness for non-devs
Sageling is a local-first AI assistant that runs entirely on Apple Silicon Macs without sending data to the cloud It uses an ~8 GB on-device model requiring at least 16 GB RAM on macOS 13 or later Designed for privacy-sensitive professionals including healthcare, education, and legal practitioners Operates without usage limits, running continuously until tasks are completed Does not train on user data; memories are stored as plain text files locally on the user's disk
Analysis
TL;DR
- Sageling is a local-first AI assistant that runs entirely on Apple Silicon Macs without sending data to the cloud
- It uses an ~8 GB on-device model requiring at least 16 GB RAM on macOS 13 or later
- Designed for privacy-sensitive professionals including healthcare, education, and legal practitioners
- Operates without usage limits, running continuously until tasks are completed
- Does not train on user data; memories are stored as plain text files locally on the user's disk
Why It Matters
Sageling represents a growing shift toward on-device AI that addresses mounting privacy and compliance concerns in regulated industries. By eliminating cloud dependency, it offers a practical solution for professionals bound by HIPAA, FERPA, and attorney-client privilege requirements without needing Business Associate Agreements. This positions local-first AI as a viable alternative to cloud-dependent tools for sensitive workflows.
Technical Details
- Runs exclusively on Apple Silicon Macs with a locally downloaded ~8 GB model; no external AI service calls unless explicitly requested by the user
- Requires macOS 13 or later and a minimum of 16 GB of unified memory to operate
- Stores short-term memories as plain text files on the local disk, fully readable, editable, and deletable by the user
- Network access is limited to web-dependent jobs, downloads, and optional user-submitted ratings—no telemetry or data uploads by default
- Uses a natural-language interface with pre-built "skills" and brief templates, targeting non-technical users rather than developers
Industry Insight
- The rise of local-first AI tools like Sageling signals increasing market demand for privacy-preserving AI, especially among regulated professionals who cannot risk data exposure
- Companies building AI assistants should consider on-device deployment options to capture enterprise and compliance-sensitive segments that cloud-only tools cannot serve
- The plain-text memory model sets a transparency standard that could pressure competitors to adopt more open, auditable data practices rather than opaque proprietary storage
Disclaimer: The above content is generated by AI and is for reference only.