Free local AI assistant for Mac that runs privately on your device
AI intelligence layer that gives AI assistants governed access to live business data
Stage
Not given
Startup
Founded
Not given
2025
Based in
Not given
US
Pricing model
Free
Free trial
Pricing
Not given
Solo $29.95/month (1 user), Team $109.95/month (5 users), Business $189.95/month (10 users). Additional users on Business plan: $9.95 each. All plans include all 40+ connectors, ChatGPT/Claude/Perplexity integration, source-cited answers, and live retrieval. Custom enterprise plans available with SS
Key features
Transcribe recordings and draft notes locally
Compare and analyze documents and spreadsheets
Run programs and access files on your Mac
Web search and research within conversations
Skills library for specific job types
Works with multiple file formats
No usage limits or subscription fees
Uses Qwen 3.5 AI model locally
Requires Apple silicon Mac with 16GB RAM minimum
AI intelligence layer for business data with cross-source reasoning across multiple connected business systems
Source-cited answers with links to the originating record
Read-only connectors with explicit safety annotations on control-plane tools
Plain English questions without dashboards required
150+ pre-built expert skills that run for your questions
Metrics validation against consistent KPI definitions
Audit-logged queries with SOC 2 alignment
Works with ChatGPT, Claude, Perplexity, and Slack
Direct MCP does not retain raw customer files or full connector response payloads
What makes it different
Safe for HIPAA, FERPA, and attorney-client privilege without agreements
Unlimited usage with no quotas or subscription required
Works offline after initial setup
Designed for non-technical users with plain language instructions
No data warehouse, no IT team, no months of setup required
Answers show where each number came from with full source attribution
Read-only by design so no changes to your connected business systems
CASA Tier 2 certified by DEKRA and hosted on Microsoft Azure
About 60% fewer tokens than AI used alone because CorpusIQ processes queries outside the chat
Sageling
CorpusIQ