Perplexity Launches Teammates AI Agents for Work

Perplexity Teammates AI agents launch

Perplexity has jumped into the growing market for persistent workplace AI agents with Teammates, a new feature letting subscribers spin up always-on digital coworkers right inside their existing Perplexity environment. The announcement puts the startup shoulder-to-shoulder with OpenAI and Anthropic, both of which rolled out comparable agent frameworks recently. Teammates are built to run continuously, adopting specialized roles within a workspace while tapping into the projects, skills, and memory already living in a user's Perplexity account.

The system leans on Perplexity Computer, the company's on-device agent runtime that executes tasks locally. According to the team, Teammates extend that foundation by adding collaboration and the capacity for ongoing, large-scale work. Each agent gets its own conversation thread inside the Perplexity app, giving you a dedicated channel to monitor progress, offer feedback, or steer the work. You can also reach those same agents through Slack, so teams can interact with them without leaving their primary communication hub.

Functionally, the offering mirrors OpenAI's Dots and Anthropic's Claude Duet. Like those systems, Teammates can be assigned distinct jobs that run in parallel-one might handle research synthesis while another drafts code or preps briefing documents. The difference, Perplexity argues, lies in the tight integration with its existing answer engine and the persistent memory layer that carries context across sessions. That continuity is meant to cut the ramp-up time that usually slows down new agent deployments.

Perplexity Max plan pricing Teammates access

Access to Teammates is gated behind the company's Max plan, which runs two hundred dollars a month. That price point sits well above the standard tiers from rivals and signals Perplexity is targeting organizations willing to pay a premium for deeper integration with its search and reasoning stack. For smaller teams or individual developers, the cost may be a barrier to experimentation, especially while competing platforms keep iterating on pricing and capability.

Perplexity CEO Aravind Srinivas framed the launch as a step toward a future where every organization runs hundreds of agents per employee. He shared that vision in a post on X, describing a workplace where digital teammates become as routine as email aliases. Whether that scale arrives depends on reliability, security, and the ability to manage fleets of agents without creating new administrative overhead.

AI workplace agents market standard layer

For people who build with or use AI, the release matters because it adds another option in a category rapidly becoming a standard layer in the modern software stack. Persistent agents that retain context, operate across tools, and collaborate with humans are moving from demos to production features. Engineers evaluating agent frameworks now have to weigh Perplexity's search-centric approach against OpenAI's broader model ecosystem and Anthropic's emphasis on constitutional AI. The choice will hinge on which platform already holds the organization's data, which pricing model fits, and which integration paths-Slack, local execution, API access-align with existing workflows.

How to enable Perplexity Teammates workspace

If you're already on Perplexity's Max plan, enable Teammates in the workspace settings and start with a single, well-scoped role like competitive research or documentation maintenance. Use the thread view to track output quality over the first week, and invite a colleague to interact with the same agent through Slack to test multi-user collaboration. Teams not on the Max tier should monitor the feature set as it evolves; Perplexity has a history of rolling capabilities down to lower plans after an initial exclusivity window. In the meantime, compare the agent behavior against OpenAI's Dots and Anthropic's Duet on identical tasks to see which system handles your domain-specific context best. If you manage procurement, factor the two-hundred-dollar monthly seat cost into any 2027 budget cycle and ask vendors for volume discounts once agent counts climb past a few dozen per team.

Frequently asked questions

What is Perplexity Teammates and how does it work?

Perplexity Teammates are persistent AI agents that run continuously inside a user's Perplexity workspace, adopting specialized roles like research synthesis or code drafting. Built on Perplexity Computer's on-device runtime, each agent gets its own conversation thread for monitoring progress and can be accessed via Slack for team collaboration.

How much does Perplexity Teammates cost and what plan is required?

Teammates access is gated behind Perplexity's Max plan at $200 per month per seat. This premium pricing targets organizations willing to pay for deep integration with Perplexity's search and reasoning stack, though it may barrier smaller teams or individual developers.

How does Perplexity Teammates differ from OpenAI's Dots and Anthropic's Claude Duet?

While functionally similar in running parallel specialized agents, Perplexity differentiates through tight integration with its answer engine and a persistent memory layer that carries context across sessions, aiming to reduce ramp-up time for new agent deployments compared to rival frameworks.

Can I use Perplexity Teammates through Slack?

Yes, Teammates can be reached through Slack, allowing teams to interact with agents without leaving their primary communication hub while maintaining the same persistent memory and conversation threads available in the Perplexity app.

What should organizations consider when evaluating Perplexity Teammates versus competitors?

Engineers should weigh Perplexity's search-centric approach against OpenAI's broader model ecosystem and Anthropic's constitutional AI emphasis. The choice hinges on which platform already holds organizational data, pricing fit, and integration paths like Slack, local execution, or API access aligning with existing workflows.

Source: The Verge