Topics
Capabilities & Architecture
Atomic actions, capability layers, architecture
What Does AI Phone Control Actually Cost? Model Choice and Token Spend
Free software does not mean zero cost: how many tokens a task uses, which operations need no model, how to pick a cheaper one, and the monthly bill.
Phone AI capabilitiesNo-codePhone Automation Without Writing Code: How Far It Actually Goes
No-code does not mean anything goes. How far phone automation without scripts reaches: which tasks plain language handles, and which still need code.
Atomic capabilitiesPhone AI capabilitiesPhone Automation Atomic Capabilities: click/swipe/input/wait_ui
Every atomic capability iEasyRun supports: launch_app, tap, swipe, input, wait_ui, album_import — how each works, its success criteria, and typical uses.
Phone AI capabilitiesAtomic capabilitiesAI Phone Control Capabilities: L1 Atomic, L2 Workflow, L3 AI
iEasyRun's three layers: atomic actions (L1), workflow templates (L2) and natural-language orchestration (L3) — what each does and how to choose.
Three-layer architectureArchitectureHow AI Phone Control Works: Inside the Three-Layer Architecture
How do vision-language models understand screens, how is intent decomposed into executable steps, and how does each atomic action guarantee correctness?