Free
Free
MIT-licensed source; hardware, developer tooling and chosen AI-provider costs are separate.
- Access
- MIT-licensed source; hardware, developer tooling and chosen AI-provider costs are separate.
Developer Tools · Official-source review 2026-10-07
iphone-use exposes a real USB-connected iPhone to AI agents through an HTTP API and MCP server. A Mac daemon runs an XCTest-based device runner, presents the screen as text, and supports tap, swipe and typing actions with reported outcomes. A browser interface and native iOS app provide human control, while saved flows can replay repeated tasks without a model.
Choose it to develop agent workflows against real iPhone apps, especially apps that lack a public API, using your own Mac and phone.
Setup requires macOS 15+, full Xcode, a development team and an unlocked USB-trusted iPhone with Developer Mode. Device-runner ports are unauthenticated; the project documents a trusted-network requirement. No device test was performed for this listing.
Read the device setup guide, prepare the trusted phone and signed XCTest runner, and validate a screen read before attempting a tap. Check whether each action was applied, not sent or unknown; save repeatable tasks as reviewed flows only after checking their outcome.
Responses distinguish applied, not-sent and unknown outcomes and indicate whether retrying is safe. Flows are reviewed per-app scripts that replay without model tokens. A successful screen action still needs a check against the app’s actual state; the directory review did not run the device runner.
Reviewed official product information on 7 October 2026. This listing describes documented workflows and billing, not a hands-on performance test. No account, device runner, generation or live customer session was used; confirm current terms before adoption.
What it can do
Unknown is different from unavailable. Each fact carries its own evidence.
| Capability | Value | Evidence | Checked |
|---|---|---|---|
| Documented operating workflow | Read the device setup guide, prepare the trusted phone and signed XCTest runner, and validate a screen read before attempting a tap. Check whether each action was applied, not sent or unknown; save repeatable tasks as reviewed flows only after checking their outcome. | Identity only | 2026-10-07 |
Understand the total cost
Free
Free
MIT-licensed source; hardware, developer tooling and chosen AI-provider costs are separate.
Conductor runs parallel coding agents on Mac with optional Conductor Cloud sandboxes, multiplayer workspaces, mobile control, and an API. It orchestrates first-party Claude Code, Codex, Cursor, and OpenCode agents using your subscriptions and keys, with isolated microVMs for long-running cloud jobs.
Explore toolCatDoes is an autonomous AI agent that builds mobile apps, websites, and related software from natural-language descriptions. It plans work, writes code, runs tests, exposes a multi-device canvas preview, supports GitHub import, and offers paid tiers with credits for deploys to web or app stores.
Explore toolCodeGPT is an AI coding agent extension for VS Code, JetBrains, and Visual Studio that supports multiple frontier models with BYOK or bundled credits. It emphasizes plan-then-build agentic workflows, MCP integrations, autocomplete, and team billing with pooled credits.
Explore toolThe practical questions
The documented setup uses a Mac daemon and Xcode to build a runner for a real iPhone. This listing does not establish support for a Windows-only or Linux-only setup.
Choose it to develop agent workflows against real iPhone apps, especially apps that lack a public API, using your own Mac and phone.
iphone-use exposes a real USB-connected iPhone to AI agents through an HTTP API and MCP server. A Mac daemon runs an XCTest-based device runner, presents the screen as text, and supports tap, swipe and typing actions with reported outcomes. A browser interface and native iOS app provide human control, while saved flows can replay repeated tasks without a model.
MIT-licensed source; hardware, developer tooling and chosen AI-provider costs are separate.. This record lists ongoing free access; check the plan limits before starting.
No paid monthly price is listed; this record treats the product as free to start. See the plan cards for entitlements, billing commitments and seat minimums.
HTTP /agent endpoints and an MCP server are documented in the official repository.. API access and subscription access may have different terms; consult the linked sources.
Setup requires macOS 15+, full Xcode, a development team and an unlocked USB-trusted iPhone with Developer Mode. Device-runner ports are unauthenticated; the project documents a trusted-network requirement. No device test was performed for this listing.
No retained pricing changes yet. A current price alone does not establish a historical trend.
Official product page reviewed for capabilities and practical constraints.
Read original source ↗MIT-licensed source; hardware, developer tooling and chosen AI-provider costs are separate.
Read original source ↗Facts apply to the named version and check date. Send a sourced correction if something changed.