> ## Documentation Index
> Fetch the complete documentation index at: https://docs.agentium.in/llms.txt
> Use this file to discover all available pages before exploring further.

# Broadcast Mode

> All agents process input; results are merged. Example with multi-perspective analysis.

# Broadcast Mode

In **Broadcast** mode, the team sends the user's input to **all** member agents. Each agent processes the task independently, and the results are merged into a single response.

***

## How It Works

<Steps>
  <Step title="User sends input">
    The user's message goes to the team.
  </Step>

  <Step title="Broadcast to all">
    Every member agent receives the same input.
  </Step>

  <Step title="Agents run in parallel">
    All agents process concurrently (where supported).
  </Step>

  <Step title="Results merged">
    The team combines outputs into one response.
  </Step>
</Steps>

***

## Example: Multi-Perspective Analysis

<CodeGroup>
  <CodeGroup.Item title="broadcast-team.ts">
    ```typescript theme={null}
    import { Team, TeamMode, Agent, openai } from "@agentium/core";

    const technicalAgent = new Agent({
      name: "Technical Analyst",
      model: openai("gpt-4o-mini"),
      instructions: "Analyze from a technical perspective: architecture, scalability, security, performance.",
    });

    const businessAgent = new Agent({
      name: "Business Analyst",
      model: openai("gpt-4o-mini"),
      instructions: "Analyze from a business perspective: ROI, market fit, user value, competitive advantage.",
    });

    const riskAgent = new Agent({
      name: "Risk Analyst",
      model: openai("gpt-4o-mini"),
      instructions: "Analyze risks: technical debt, compliance, operational, and strategic risks.",
    });

    const analysisTeam = new Team({
      name: "360° Analysis",
      mode: TeamMode.Broadcast,
      model: openai("gpt-4o"),
      members: [technicalAgent, businessAgent, riskAgent],
      instructions: `
        Each member analyzes the input from their perspective.
        Merge all analyses into a comprehensive report with sections:
        1. Technical Analysis
        2. Business Analysis
        3. Risk Analysis
        4. Summary & Recommendations
      `,
    });

    const output = await analysisTeam.run(
      "Evaluate adopting microservices for our monolithic e-commerce platform."
    );
    console.log(output.text);
    ```
  </CodeGroup.Item>
</CodeGroup>

***

## When to Use Broadcast

<CardGroup cols={2}>
  <Card title="Multi-perspective" icon="eye">
    Get technical, business, and risk views in one pass.
  </Card>

  <Card title="Ensemble answers" icon="layer-group">
    Combine multiple expert opinions for robustness.
  </Card>

  <Card title="Parallel processing" icon="bolt">
    All agents run concurrently for faster results.
  </Card>

  <Card title="Comprehensive coverage" icon="check-double">
    Ensure no angle is missed.
  </Card>
</CardGroup>

***

## Considerations

<Accordion title="Cost and latency">
  Broadcast invokes every member, so cost and latency scale with team size. Use smaller/faster models for members when appropriate.
</Accordion>

<Accordion title="Merge strategy">
  The orchestrator model merges results. Clear instructions help it synthesize a coherent final output.
</Accordion>