---
id: "claim-two-diverse-beats-sixteen"
type: "claim"
source_timestamps: ["¶4"]
tags: ["performance-metrics", "efficiency"]
related: ["claim-diversity-improves-performance", "quote-two-beats-sixteen"]
confidence: "high"
testable: true
speakers: ["Mark Purdy"]
sources: ["agentic"]
sourceVaultSlug: "hbr-seg-agentic"
originDay: 6
articleStem: "hbr-new-28-agent-teams-different-models"
sourceUrl: "https://hbr.org/2026/06/the-strongest-teams-of-ai-agents-will-be-built-using-different-models"
sourceTitle: "The Strongest Teams of AI Agents Will Be Built Using Different Models"
---
# Two diverse agents can match or exceed the performance of 16 homogeneous agents

**Claim:** A cited study shows that a minimal team of **just two diverse agents can match or even exceed** the output and problem-solving capability of a massive team of **16 homogeneous agents** (see the verbatim [[quote-two-beats-sixteen]]).

**Implication:** Scaling up the *number* of agents without scaling their *diversity* yields rapidly diminishing returns — diversity, not headcount, is the lever. This is the sharpest quantitative expression of the article's thesis and directly supports [[claim-diversity-improves-performance]].

**Confidence: high** (as stated) — but see caveat.

**Enrichment assessment — CASE STUDY, NOT A LAW:** The 2-vs-16 figure appears to come from a single experiment rather than a broad meta-analysis, and it does not match any widely cited named study. Agent performance is highly domain-dependent — in some tasks larger homogeneous ensembles or a single strong model may outperform a small heterogeneous team. Interpret this as an **illustrative case-study result** that warrants replication and domain-specific evaluation before extrapolation.