Match an image, then click
Find a specified button, icon, or image on screen and click it automatically. The Flow follows the matched target even when its position changes.
LaiCai Flow in LaiCai Screen Mirroring
Android automation with AI
Open apps, find a button or text on screen, and repeat the steps you need. LaiCai Flow is the automation feature in LaiCai Screen Mirroring: build a saved workflow visually, with AI assistance when your task needs it.
No programming requiredDrag-and-drop nodesAI model supportAI Planner and MCPReusable Flow Profiles
Get started
Build an Android task you can run again
Start with a small task, such as opening an app, finding a visible button, tapping it, and checking the next screen.
- 01
Install LaiCai Screen Mirroring on Windows or macOS, connect an Android phone or supported emulator, and check that you can see and control its screen.
- 02
Open LaiCai Flow and create a workflow. Drag in the actions you need, or use AI Planner or MCP to help prepare a draft from your task description.
- 03
Choose how to find the target: a saved button image, text recognition, an Android UI element, or a configured detection model. Capture or configure the inputs your task needs.
- 04
Connect the action and the result check. Decide what should happen if the target is missing, so the task does not continue with an unrelated tap.
- 05
Run the task on one device. Inspect the actual screen, screenshots, node results, and run logs to confirm that it reached the expected result.
- 06
Save the Flow Profile and rerun it from the same starting state. Adjust inputs and screen checks before reusing the workflow on other phones or emulator instances.
Workflow
Automate actions that depend on the screen
Read and analyze screen text
Read page messages, order status, form content, or search results, then use the text to decide, classify, or continue to the next step.
Repeat fixed actions
Connect actions such as opening an app, tapping, typing, swiping, and waiting into a Flow, then save it and run the same steps whenever needed.
Android app automation testing
Run UI steps automatically, verify buttons, text, and results, then keep screenshots and run logs for app testing and regression testing.
Automation comparison
Choose an Android automation tool for your task
| Your task | LaiCai Flow | AI agents | Macros and scripts | Test frameworks |
|---|---|---|---|---|
| Create the task | Connect visual nodes; AI Planner or MCP can help configure them. | Describe a task; the agent's supported tools determine what it can do. | Record inputs or write the actions and conditions. | Write and maintain tests using the framework's device controls. |
| Check the screen | Use image matching, OCR, UI elements, or a configured AI model. | Screen and tool access depend on the agent. | Fixed replay repeats input; scripts can add checks when implemented. | Element assertions and image checks depend on the framework and setup. |
| Repeat and inspect | Save a Flow Profile, reuse inputs, and inspect run logs and screenshots. | Scheduling, repeatability, and logs vary by product. | Reuse recordings or code; update them when the app changes. | Reuse test suites and inspect test reports. |
| A useful starting point | A repeatable Android task you want to build visually. | A task that benefits from model-driven decisions. | A stable sequence or a workflow you prefer to maintain in code. | An engineering team building an automated test suite. |
FAQ
Android automation tool FAQ
What is LaiCai Flow?
LaiCai Flow is an Android automation tool inside LaiCai Screen Mirroring. Build a saved workflow by connecting visual nodes for actions, screen checks, conditions, and task data. It supports AI-assisted setup and optional AI models inside a running workflow.
Do I need programming skills or an AI model for every task?
You can create flows by dragging and connecting nodes without writing code. AI Planner and MCP can help configure a draft, while runtime AI nodes are a separate choice. Basic actions and template matching do not need a runtime LLM; OCR, detection, or LLM steps need their corresponding engine or model configured.
Can it find a button, read text, or recognize an object?
Yes. Use a saved image for template matching, OCR to read a screen region, or Android UI elements when the app exposes them. For object detection, configure a supported local YOLO/ONNX model. Check the result on your actual app before connecting it to a tap.
See recognition and action setupCan an AI model analyze screen images and text?
Yes. Text analysis can use OCR, API, file, or earlier step results. A vision-capable model can analyze a captured screen image and return structured data for a later decision. Configure the model and inspect its output; a plausible response alone does not prove that the phone reached the correct screen.
What actions can I automate on an Android phone?
Flows can tap, swipe, long-press, drag, type text, send keys, use the clipboard, open or stop apps, and go Home or Back. Combine these with waits, conditions, loops, and reusable child flows. HTTP requests, ADB commands, AutoJS scripts, saved macros, files, and stored data can also be part of a workflow.
Can I run the same task on multiple phones or emulators?
You can reuse a saved Flow Profile across supported Android targets. First check the task on one device, then review app state, screen layout, assets, and inputs on each additional target. A flow that works on one screen is not automatically verified on every phone.
Organize multiple Android phonesDoes the computer need to stay connected?
Desktop LaiCai Flow runs its logic and image analysis on the computer and sends actions to the connected phone. For compatible workflows that need to run on the phone itself, use LaiCai Flow Inside and the LaiCai Android Agent. Network-dependent steps still require network access.
Read the Flow Inside deployment guideCan the workflow speak a result or voice reminder?
Yes. The speech action uses the local speech engine on the desktop LaiCai computer. It can read a reminder or a result aloud there; it does not make the Android phone speak.