PrompteryxBeta
    ⌘K
    PlatformSupport
    Menu
    IntroductionInstall Desktop & extensionQuick StartPlatform Overview
    OverviewWhat you can makeAI Browser Agent in the chatModels and AutoConnectorsSchedulesProjectsExplore and publishingShare linksMemory and SkillsPersonal agent (beta)Welcome, settings and skinsPlans, credits and top-upsPrompteryx DesktopGroups (pilot)Phone app (early access)
    OverviewCreating WorkflowsTesting & DebuggingScheduling
    OverviewGetting StartedCommandsRecordingSettingsStructured OutputParallel AgentsTips & Best Practices
    OverviewInstallationSidebar AgentUsing Attached FilesExecute on Current TabExporting to Visual Studio
    Overview
    OverviewOpen PageClickClick CoordinatesFillExtract TextScreenshotWaitPress KeyFocus ElementBlur ElementSelect TextHighlight ElementSet ViewportGet Bounding BoxScrollHoverSelect OptionAssertionsCookiesEvaluate JavaScriptExample Workflows
    OverviewPrompteryx Browser AgentOpenAI
    OverviewSchedule TriggerWebhook Trigger
    OverviewGoogle SheetsGoogle CalendarConnect Hub (3,000+ apps)Example Workflows
    AI Vision SettingsChrome Automation ProfilesCloud BrowserBrowser Profiles & Persistent ContextVideo RecordingExecution LogsSharing & CollaborationEdit History
    Selectors GuideVariables
    Start hereGet an API keyRun the AI browser agentDrive a cloud browserRun workflows from anywhereUse connected apps (Connect Hub)SDK (TypeScript)CLIMCP for AI assistants
    REST APIWebhooksMCP toolsCloud Browser APICloud Browser quickstartMigrate from Browserbasen8nZapierMake.com

    Need help? Contact Support

    1. Documentation
    2. Nodes
    3. Control Flow
    4. Parallel

    Parallel Execution

    Control Flow - Advanced

    Run multiple branches simultaneously for faster workflows

    Overview

    Parallel execution allows you to run multiple branches of your workflow at the same time. This is particularly powerful for browser automation where you can process multiple tabs or pages simultaneously, dramatically reducing total execution time.

    Browser Automation
    In browser workflows, each parallel branch operates in its own tab, allowing you to perform actions on multiple pages at once.
    The Parallel Split node's config panel: number of branches, wait for all branches toggle, continue on error toggle
    The Parallel Split config panel — set the branch count and whether the join waits for every branch.

    Parallel Nodes

    Parallel Split
    Start parallel execution by creating multiple branches
    Parameters
    Configuration options for this node
    branches
    Number
    Required

    Number of Branches: how many parallel branches to run (each in its own tab), 2–10

    Default: 2
    waitForAll
    Boolean

    Wait for All Branches: wait for every branch to finish; turn off to continue once the first branch completes

    Default: true
    continueOnError
    Boolean

    Continue on Error: keep going if some branches fail; when off, a failing branch stops the run

    Default: false

    Each connection you draw out of the Parallel Split starts a branch. A branch runs until it reaches the Parallel Join; branch steps are numbered 2A.1, 2B.1, and so on.

    Parallel Join
    Merge parallel branches back into a single flow
    Parameters
    Configuration options for this node
    mergeStrategy
    Select
    Required

    Merge Strategy: Collect All Results (array), Use First Result Only, or Use Last Result Only

    Default: Collect All Results (array)
    outputVariable
    String
    Required

    Output Variable Name: where the merged results are stored

    Default: parallelResults

    With the default settings, results from each branch are available in {{parallelResults}} as an array — one entry per branch with its success flag, data and any error.

    Usage Examples

    Example 1: Scrape Multiple Pages
    Extract data from 3 pages simultaneously

    Workflow structure:

    1. Parallel Split (3 branches)

    Branch 1: Open Page → example.com/page1 → Extract data

    Branch 2: Open Page → example.com/page2 → Extract data

    Branch 3: Open Page → example.com/page3 → Extract data

    2. Parallel Join (collect all results)

    3. Combine all extracted data

    4. Save to Google Sheets

    ✅ The three pages are processed at the same time instead of one after another.

    Example 2: Monitor Multiple Sites
    Check availability across different websites

    Workflow structure:

    1. Parallel Split (branches for each site)

    Branch 1: Check Amazon price

    Branch 2: Check eBay price

    Branch 3: Check Walmart price

    2. Parallel Join

    3. Compare prices and find best deal

    4. Send notification with results

    Example 3: Fill Forms in Parallel
    Submit multiple forms at once

    Workflow structure:

    1. Parallel Split (2 branches)

    Branch 1: Fill registration form on Site A

    Branch 2: Fill registration form on Site B

    2. Parallel Join

    3. Log completion status for both

    How Tabs Work

    When using Parallel Split in browser workflows:

    • Each branch operates in its own browser tab
    • Tabs are created automatically when branches start
    • Actions in one branch don't affect other branches
    • The Parallel Join merges the branches' results into its output variable
    • Branch logs are tagged so you can filter them by branch in Execution Results

    Best Practices

    Limit Branch Count
    More branches = more browser tabs = more memory. Keep parallel branches to 3-5 for optimal performance. Too many tabs can slow down execution.
    Handle Branch Failures
    If one branch fails, decide whether to continue with others or fail the whole workflow. Use the Parallel Split's "Continue on Error" and "Wait for All Branches" settings accordingly.
    Name Your Branch Steps
    Give the nodes in each branch descriptive names, and use the path filter (Branch-1, Branch-2, ...) in Execution Results to read one branch's logs at a time.
    Avoid Shared State
    Don't rely on variables set in one branch being available in another during parallel execution. Each branch should be independent until they rejoin.

    Parallel vs For Each

    Use Parallel Split when:

    • You have different actions for each branch
    • Branches have different destinations/logic
    • You need maximum speed

    Use For Each when:

    • You're doing the same thing to many items
    • Order matters
    • You have many items (10+) to process

    Related Documentation

    For Each Loop

    Sequential iteration

    Testing Workflows

    Debug parallel execution