A Mac app that lets AI agents control a real iPhone
Eight specialized AI agents research your question in parallel and deliver one clear answer you can act on
Stage
Not given
Startup
Founded
Not given
2024
Based in
Not given
SI
Pricing model
Paid
Freemium
Pricing
Monthly: $25/month for 1 iPhone. Lifetime: $99 once for 1 iPhone. Lifetime Plus: $199 once for up to 5 iPhones. All plans include access to Claude Code and Codex on your own subscription, run on your Mac with no cloud, and include all updates.
Not given
Key features
Control a real iPhone from your Mac over USB and Bluetooth
Connect Claude Code, Codex, Cursor, Hermes, OpenClaw and any MCP client
Built-in chat that runs on your own Claude or ChatGPT subscription
Use any App Store app and your own logged-in accounts
No installation on the iPhone, no developer mode needed
No Xcode, Apple Developer account or code signing required
Optional phone unlock with passcode stored in macOS Keychain
Local MCP server that only accepts connections from your Mac
Free trial for the first hour of live use
Works with up to 3 Macs per license
Eight parallel specialist agents analyzing different business data domains
Stripe and Google Ads revenue tracking with MRR and CAC analytics
Google Analytics 4 and Search Console traffic analysis
PostHog product analytics and funnel monitoring
Competitor website analysis and pricing extraction at scale
Real-time chat workspace with cited sources for all findings
OAuth secure, read-only data connections
Signal boards for ongoing metric monitoring
What makes it different
Real iPhone instead of simulator - runs your actual apps and accounts with no virtual copy limitations
Local-only operation with no account or cloud - your data stays on your Mac
No developer setup required - just plug in a USB cable and pair Bluetooth
Bring your own AI - works with agents and plans you already pay for
Pay once, keep it for life with no subscriptions required
Eight specialized agents work in parallel instead of a single chatbot
Every claim cited to live sources, not generated from training data
Bootstrapped with roadmap driven by customer needs, not investor targets
Data analyzed and discarded, never trained on models
Zeptap
Inteldo