Teams of AI agents barely outperform solo agents but cost up to 5.1x more, according to Vals AI. Only one out of four tests with GPT-6 Sol and Claude Opus 5.5 showed a measurable gain. Anthropic's own data backs this up: beyond ten agents, quality plateaus while token costs keep climbing. The article AI agent teams waste massive tokens for barely measurable quality gains, research finds appeared first on The Decoder.
Editorial read: "AI agent teams waste m..." is a model capability and product roadmap update; source signal: The Decoder as a news portal, AI 79, still useful as recent context; value: tracking model iteration and competitive response.