D1 · Agentic Architecture › Cluster 4 · Multi-Agent

4.1 · Single-Agent-+-Tools vs. Multi-Agent

Pick the simplest tier that works

CONCEPT The escalation, one tier up

4.1 lives inside the agent tier from 1.1. Once you've decided you need an agent, the next restraint question is: one agent with a rich toolbox, or several agents coordinating? The architect's default is unchanged — pick the simplest tier that solves the problem. A single agent + tools is simpler, cheaper, and lower-latency; multi-agent is a deliberate escalation you justify, not a starting point.

DimensionSingle agent + toolsMulti-agent
StructureOne model loop, many tools (hub-and-spoke)A coordinator plus specialist sub-agents, each with its own context
Cost / latencyLower — one context, fewer round-tripsHigher — extra agents each cost tokens; sequential ones add latency
PredictabilityEasier to test and reason aboutMore moving parts, more failure modes
Best fitMost agent tasksGenuinely parallel sub-tasks, 5+ step workflows, distinct specialised roles
🔑 Default to a single agent + tools. Reach for multi-agent only when the task is genuinely parallel, is a 5+ step workflow, or needs distinct specialised roles. Cost- or latency-sensitive? Stay single.

CONCEPT The three "go multi-agent" triggers

Multi-agent earns its overhead only when at least one of these clearly holds:

⚠️ None of these hold? A single agent + tools is the design. If you're cost- or latency-sensitive, that pressure points firmly at single-agent.

DIAGRAM The decision tree

flowchart TD A([You've decided you need an agent]) --> Q1{"Genuinely parallel
independent sub-tasks?"} Q1 -->|Yes| M["🟣 Multi-agent
parallel fan-out"] Q1 -->|No| Q2{"5+ step workflow
bloating one context?"} Q2 -->|Yes| M2["🟣 Multi-agent
pipeline / coordinator"] Q2 -->|No| Q3{"Distinct specialised
roles needed?"} Q3 -->|Yes| M3["🟣 Multi-agent
coordinator + specialists"] Q3 -->|No| S["🟢 Single agent + tools
hub-and-spoke · cheapest"] Q4{"Cost / latency
sensitive?"} -.forces.-> S classDef s fill:#d4edda,stroke:#28a745,color:#155724; classDef m fill:#e2d4f0,stroke:#764ba2,color:#3d1a5b; class S s; class M m; class M2 m; class M3 m;

TRADE-OFFS What multi-agent buys and costs

✅ Multi-agent gains

  • Parallelism cuts wall-clock on independent work
  • Specialised prompts/tools/models per role
  • Context isolation keeps each sub-task focused
  • Decomposes 5+ step workflows into digestible units

⚠️ Multi-agent costs

  • More tokens — every agent has its own context
  • Sequential agents add latency additively
  • State must be passed explicitly (see 4.5)
  • More failure modes; harder to test
Rule of thumb: single agent + tools → multi-agent, and escalate only when a concrete trigger (parallel / 5+ steps / distinct roles) forces it.

EXAM TRAPS

Trap: Reaching for multi-agent because the task "sounds complex." → Complexity alone doesn't justify it. Look for a concrete trigger: genuine parallelism, 5+ steps, or distinct roles. Absent those, single agent + tools wins.
Trap: Ignoring cost/latency signals. → Words like "cost-effective," "low-latency," or "high-volume" in the stem steer you toward a single agent, not away.
Trap: Assuming two sequential specialists are "parallel." → Parallelism only helps when sub-tasks are independent. Dependent steps run sequentially and add latency (see 4.6).
Trap: Thinking multi-agent is "more powerful, so better." → It's an escalation with real overhead. The exam rewards the simplest tier that works.

PRACTICE

Q1 · A support bot needs to look up orders, check inventory, and issue refunds — sequentially, per conversation. Single or multi-agent?

Single agent + tools. The work is sequential and dependent, not parallel, and the three actions are just tools, not distinct specialised roles. One hub-and-spoke agent with three tools is simpler, cheaper, and lower-latency.

Q2 · A research report must gather facts on 8 independent companies, then synthesize. Single or multi-agent?

Multi-agent. The 8 lookups are genuinely independent, so parallel fan-out sub-agents cut wall-clock, and a coordinator synthesizes. This hits the "genuinely parallel sub-tasks" trigger.

Q3 · A stakeholder says "make it as cheap and fast as possible." How does that bias the single-vs-multi call?

Toward single agent + tools. Multi-agent adds tokens and (if sequential) latency. Only override that bias if a hard trigger — true parallelism or genuinely distinct roles — is present.