GEELARK MCP

Connect AI tools to GeeLark with MCP

Manage cloud phones, browser profiles, and supported automation tasks through conversations with MCP-compatible AI tools such as Codex, Claude, and Cursor.

GeeLark MCP is a Model Context Protocol server that connects compatible AI tools to GeeLark’s cloud phone, browser profile, and automation capabilities.

GeeLark MCP connecting an AI client to cloud phone and browser profile management

FROM REQUEST TO ACTION

Turn AI requests into GeeLark operations

Describe the task in everyday language. Your connected AI tool can use available GeeLark tools to retrieve information and perform supported operations.

Organize profiles and resources

Review groups, tags, and proxies, then use supported tools to apply the changes you choose.

CONNECTED CAPABILITIES

Manage cloud phones, browser profiles, and tasks

Available tools, permissions, and runtime requirements determine what your AI tool can do. Cloud phone and browser operations may have different requirements.

Explore cloud phones →

Explore multi-account browser →


AreaSupported examples
Cloud phonesFind, create, edit, start, stop, and inspect supported cloud phone profiles.
Browser profilesCreate, manage, and launch browser profiles, with supported browser task operations.
AutomationCreate or schedule supported tasks and review task history.
OrganizationWork with available groups, tags, and proxy-management tools.
Apps and filesUse supported app and file operations for cloud phone workflows.

A PRACTICAL EXAMPLE

Understand your cloud phone setup with one request

The same connection can support other available operations. What happens next depends on your request, the tools exposed to your AI client, and your GeeLark access.

Example workflow showing an AI request, a GeeLark MCP tool call, and a cloud phone summary by Android version

Ask your AI tool to retrieve selected cloud phones and summarize them by Android version. GeeLark MCP provides the device information, and your AI tool organizes the results.

HOW IT WORKS

Connect your AI tool to GeeLark operations

01

Connect your AI client

Configure the official GeeLark MCP server in an MCP-compatible AI tool.

02

Describe your task

Your AI tool interprets your request and selects from the available GeeLark tools.

03

Review the result

GeeLark performs the supported operation and returns information your AI tool can present.

GeeLark MCP can run through the official geelark-cli package in MCP server mode. API, CLI, and MCP offer related ways to access supported GeeLark capabilities.

THREE WAYS TO CONNECT

Choose API, CLI, or MCP

Find all three in GeeHub. Choose the interface that fits how you work.

MethodHow you access GeeLarkCommon starting point
APIRequests from your application or systemBest for custom integrations
CLICommands from your terminal or scriptsBest for repeatable command-line operations
MCP (this page)Tool calls from a compatible AI clientBest for conversational access to supported operations

These are common starting points, not exclusive capabilities. Technical users and AI agents may combine methods.

GET CONNECTED

Set up your GeeLark MCP connection

02

Configure your AI tool

Follow the official instructions for your MCP-compatible client and configure authentication locally.

Keep authentication tokens private. Check the documentation for installation and local browser connection requirements.

FAQs

GeeLark MCP connects compatible AI clients to supported GeeLark operations using Model Context Protocol. Your AI tool can call available tools to work with cloud phones, browser profiles, and other supported resources.
MCP-compatible tools such as Codex, Claude, and Cursor can connect when configured according to their requirements. See the setup guide for current instructions.

Yes. GeeLark MCP provides supported operations for both. The available actions and connection requirements can differ between cloud phones and browser profiles.

It supports available task-management operations, including creating or scheduling supported tasks and reviewing task history. It does not mean every possible workflow can be generated or executed automatically.

API connects your applications and systems through programmatic requests. CLI provides terminal commands. MCP makes supported operations available as tools inside compatible AI clients.

No. GeeLark MCP connects external AI tools to GeeLark operations. GeeLark AI provides content-creation tools such as image and video generation.

You need GeeLark API access, authentication configured locally, a compatible AI client, and the runtime required by your chosen setup method.

The AI client and MCP server must be available when making tool calls. Browser operations can depend on a local connection. Requirements for tasks already submitted to GeeLark depend on the task type; check the documentation.

Bring GeeLark into your AI workflow

Connect your AI tool to cloud phones, browser profiles, and supported GeeLark operations.