Perplexity has launched "Teammates," always-on AI agents designed for ongoing, large-scale work, positioning the search company directly against OpenAI's Dots and Anthropic's Claude Duet in the fast-growing market for persistent digital coworkers.
The launch, reported by The Verge on Thursday, lets subscribers assign multiple AI Teammates different jobs that run simultaneously around the clock — a shift from chatbots that answer questions on demand toward agents that hold responsibilities over days and weeks. For more context on this story, see our ongoing more AI stories.
What Teammates Actually Do
Each Teammate can take on multiple roles within a workspace, pulling existing projects, skills and memory from the Perplexity tools a customer already uses. That memory carry-over matters: instead of briefing a fresh agent from scratch, Teammates inherit the context accumulated in a team's Perplexity workspace.
Teammates build on Perplexity Computer, the company's AI system that deploys agents on a customer's device to complete tasks. Perplexity says the new agents extend Computer by adding the ability to collaborate and to carry out "ongoing, large-scale work" — the kind of multi-step projects that outlast a single chat session.
Computer itself has been evolving quickly. In August, Perplexity launched its Portable Computer, which the company says runs "entirely on device," and by late October The Verge reported that Perplexity Computer can now run fully offline on Mac — provided the machine has enough RAM. Teammates layer persistent, role-based coordination on top of that device-level agent foundation.
Every Teammate gets its own thread, so progress is tracked individually rather than buried in a shared conversation. Users can monitor each agent through threads in the Perplexity app, and — critically for workplace adoption — interact with their Teammates inside Slack, where much of office coordination already lives.
"Over time, we envision every org having hundreds of teammates working for each person," Perplexity CEO Aravind Srinivas said in a post on X announcing the feature.
The Price of a Digital Coworker
There is a catch for would-be employers of AI staff: Teammates require a subscription to Perplexity's $200-per-month Max plan, the company's top tier. The Verge notes that this makes Perplexity's take on persistent agents considerably more expensive than rival offerings — OpenAI's Dots and Claude Duet both target lower price points for comparable role-based agent work.
The premium pricing signals Perplexity's strategy: rather than compete on cost, the company is betting that teams will pay enterprise-style rates for agents that stay continuously engaged with their projects and tools.
The math behind that bet is straightforward. A salaried employee costs an organization tens of thousands of dollars a year; a $200 monthly subscription costs $2,400. If a Teammate reliably handles even a fraction of a knowledge worker's routine workload — compiling research, monitoring projects, drafting reports — the subscription pays for itself. The risk is equally clear: agents that stall, hallucinate or need constant supervision convert that arithmetic into an expensive distraction.
From Answer Engine to Agent Platform
The launch also marks another step in Perplexity's transformation from an AI-powered answer engine into a broader work platform. The company built its reputation on search with citations, then expanded into a browser, an assistant and device-level agents. Teammates represent the logical endpoint of that trajectory: if Perplexity's core competency is finding and synthesizing information, then selling always-on agents that do that finding and synthesizing continuously is a natural product evolution.
It is also defensive. As OpenAI and Anthropic push agents that can occupy the same workspace territory, Perplexity cannot afford to cede the "AI coworker" category to rivals whose models power competing search products. The company's Amazon-related legal battles over its shopping agent, which have already reached the appeals court, underline how contentious agent territory has become.
A Crowded Race for the AI Coworker
The major AI labs have converged on the same product concept this year — named, tellingly, with the same metaphor. OpenAI's Dots and Anthropic's Claude Duet both frame AI agents as colleagues that hold jobs, rather than assistants that answer prompts. Perplexity's Teammates is the third major entrant, and the first to lean so heavily on integration with an existing search-and-research stack.
The differentiation for Perplexity lies in its heritage. The company built its name on AI-powered search with citations, and Teammates inherit that research orientation: agents that can investigate, compile and report continuously fit naturally alongside a search engine's existing workflow.
Why It Matters
The launch underscores how quickly the AI industry has moved from answering questions to doing work. Persistent agents that operate 24/7 raise practical questions — about oversight, error correction and accountability — that one-shot chat sessions never had to answer. They also put pressure on the subscription economics of every AI company: if a $200-per-month agent genuinely replaces hours of knowledge work, the price looks cheap; if it needs constant supervision, it looks like a novelty.
For now, Perplexity is asking its most committed users to make that bet. Whether hundreds of digital teammates per organization becomes the norm, as Srinivas predicts, or an expensive experiment will depend on how these agents perform when the work is real.
Stay Ahead of AI
For more on the latest AI tools, agents and model launches, follow AI Buzz Wire.
Read more AI news at aibuzzwire.news.
