Featured image courtesy of Claude.

Perplexity Computer just added Multiagent, a new system that runs up to 16 AI agents to share findings and cross-check work. This advanced feature, first reported by MarkTechPost, draws on models from Claude, OpenAI’s GPT, and Google’s Gemini. Claude’s models are now in a complex, multi-agent system. This is a significant development for Claude. It positions Claude as a foundational technology provider, capable of powering sophisticated AI applications beyond simple conversational interfaces.

This news comes as Apple, according to TechCrunch, appears to be developing its own AI agent designed to run on macOS. Apple wants to compete with Meta Muse, Instinct, OpenAI Dots, and Claude. For Claude, being named alongside these established players by a company like Apple shows its growing prominence and competitive standing in the emerging personal AI agent market. Claude is now a direct challenger for end-user AI experiences, not merely a backend provider.

Claude’s Path: From Safety Discussions to Core Infrastructure

These recent developments expand Claude’s public profile beyond its earlier focus on AI safety and provenance. Just three months ago, in August 2026, this outlet covered “Claude AI Watermarks: What They Mean for Indian Startups and Global Regulation.” At that time, the discussion centered on Claude’s involvement in embedding digital watermarks into AI-generated content. That effort showed a commitment to responsible AI development and addressed concerns about misinformation and intellectual property. It also addressed broader regulatory discussions around AI content identification.

The shift, or rather the expansion, is clear. While the watermarking initiative positioned Claude as a proponent of AI safety and transparency, the company is now simultaneously demonstrating its capabilities as a core AI engine for complex tasks and a direct competitor in the high-stakes AI agent race. Its models perform well enough to be part of Perplexity’s “Ultra” effort level, where a lead agent manages up to 16 subordinate agents. This suggests a level of reliability and performance that makes Claude suitable for demanding computational environments.

Agent systems themselves represent a significant evolution in AI. They move beyond static models to autonomous entities capable of planning, executing, and iterating on tasks. For Apple to specifically call out Claude as a competitor in this area indicates that Claude’s agent-focused research and development are keeping pace with, or even leading, industry giants.

Implications for the AI Market

The dual role Claude now occupies, as both a foundational model provider and a direct competitor in the agent space, hints at a broader strategy. By making its models available through APIs for services like Perplexity, Claude increases its footprint across the AI industry. Simultaneously, by developing its own user-facing agent experiences, it aims for direct market share. This approach allows Claude to influence the industry at multiple levels.

Competition for AI agents is quickly solidifying, with major players like Meta, OpenAI, and now Apple making their moves. Claude’s presence in this list, alongside its models powering multi-agent systems, shows its position as a central figure in the AI arms race. The market will watch to see how Claude balances its commitment to AI safety, shown by its watermarking work, with the aggressive pursuit of capability and market share.

Quick Facts

What is Perplexity’s new Multiagent feature?

Perplexity Computer now includes Multiagent, a system that allows a lead agent to run up to 16 AI agents that share findings and cross-check each other’s work.

Which AI models does Perplexity Multiagent use?

Perplexity Multiagent uses models from Claude, OpenAI’s GPT, and Google’s Gemini.

Who does Apple see as competitors for its AI agent?

Apple is developing AI agents to compete with Meta Muse, Instinct, Claude, and OpenAI Dots.

Compiled by Launch91 Desk from the sources linked above. More about Launch91.