Back to list
Perplexity Launches Always-On AI Teammates to Challenge OpenAI and Anthropic in Workspace Automation
Product LaunchPerplexityAI AgentsEnterprise AI

Perplexity Launches Always-On AI Teammates to Challenge OpenAI and Anthropic in Workspace Automation

AI startup Perplexity has officially entered the workplace agent race by introducing "Teammates," a suite of always-on artificial intelligence agents engineered to handle diverse organizational tasks. Designed to compete directly with enterprise solutions from OpenAI and Anthropic, Teammates can fulfill multiple workspace roles by drawing on users' existing projects, learned skills, and persistent memory across the Perplexity platform. Crucially, the system works in tandem with Perplexity Computer, an underlying AI system capable of deploying agents directly on a user's device. This release marks an ambitious transition for Perplexity, expanding its core capabilities from an answer-based search engine into an active, autonomous workforce ecosystem built to execute complex, context-rich productivity workflows across modern enterprise environments.

The Verge

Key Takeaways

  • Direct Competitive Challenge: Perplexity has launched work-focused artificial intelligence agents named "Teammates," positioning itself in direct competition with frontier AI leaders OpenAI and Anthropic.
  • Always-On Workspace Multi-Tasking: Unlike reactive prompt-based chatbots, Teammates function as persistent, always-on digital workers capable of assuming multiple defined roles across enterprise workspaces.
  • Integrated Context and Memory: Teammates natively incorporate users' existing Perplexity workflows, pulling from active projects, configured skills, and historical memory to preserve continuity across tasks.
  • On-Device Deployment via Perplexity Computer: The new agents integrate tightly with Perplexity Computer, an AI architecture designed to deploy agents directly on local user devices.
  • Strategic Shift to Autonomous Workflows: This release signifies Perplexity's strategic transition beyond conversational information retrieval and search into autonomous workplace collaboration.

In-Depth Analysis

The Shift Toward Persistent, Always-On Workplace Agents

The introduction of Teammates represents a pivotal evolution in how generative artificial intelligence is operationalized inside business environments. For much of the generative AI boom, enterprise adoption was centered around passive, conversational interfaces where human workers initiate every single transaction through discrete prompts. Perplexity's introduction of "always-on" agents breaks with this prompt-and-wait model. By designing agents that remain active continuously within a workspace, Perplexity is shifting the operational paradigm toward proactive collaboration.

In a typical organizational setting, tasks are rarely isolated, one-off interactions; they require continuity, routine follow-ups, and adaptability across changing priorities. According to the original report, Teammates are explicitly designed to adopt multiple roles within a workspace. Rather than requiring users to spin up new chat sessions or manually establish context for different domains, a single Teammate architecture can navigate diverse responsibilities. This persistent operational posture allows digital coworkers to oversee multi-stage workflows without relying on continual re-prompting from human operators.

Contextual Continuity: Integrating Projects, Skills, and Memory

A critical limitation of first-generation workplace AI implementations has been context fragmentation. When an AI system forgets previous tasks or operates in isolation from the files and knowledge bases an employee uses daily, the human overhead of guiding the tool often outweighs the productivity gains. Perplexity directly addresses this pain point by grounding Teammates in the tools and data users have already accumulated.

Teammates can systematically pull from existing projects, customized skills, and long-term memory previously stored across Perplexity tools. This capability ensures that the agents do not operate as blank slates. By maintaining awareness of prior deliverables, operational guidelines, and organizational history, Teammates can ground their activities in established corporate context. Coupling persistent memory with domain-specific skills bridges the gap between high-level reasoning models and execution-oriented enterprise workflows, allowing an agent to step into a workspace with the background knowledge of an existing team contributor.

On-Device Deployment and Perplexity Computer

A distinguishing technical pillar of this launch is the integration of Teammates with Perplexity Computer. Described as an AI system capable of deploying agents directly on a user's device, Perplexity Computer provides the execution environment necessary for real-world workplace utility. Web-confined AI tools frequently encounter friction when interacting with local software ecosystems, native file structures, and secure internal tools.

By facilitating on-device agent deployment, Perplexity Computer enables Teammates to interact with a user's actual workstation environment. This architecture allows AI agents to interface directly with local applications, perform complex actions within desktop workflows, and complete tasks that require native computing capabilities. While specific technical parameters regarding local hardware utilization and sandboxing protocols remain to be expanded in subsequent documentation, the integration clearly signals Perplexity's intent to move AI execution out of disconnected cloud silos and directly onto the employee's operational machine.

Industry Impact

Escalating Enterprise Rivalry with OpenAI and Anthropic

The launch of Teammates marks a major strategic maneuver by Perplexity to contest enterprise market share against industry heavyweights OpenAI and Anthropic. Both competitors have placed massive strategic emphasis on autonomous agents and computer-use capabilities. OpenAI has steadily integrated agentic features and custom workspaces, while Anthropic has pioneered direct desktop interaction with its computer-use models. By deploying Teammates through Perplexity Computer, Perplexity demonstrates that it no longer views itself purely as an AI-powered search alternative, but rather as an end-to-end enterprise platform.

This rivalry is accelerating the timetable for autonomous workplace technologies. As frontier AI labs compete to offer persistent digital labor rather than isolated conversational assistants, enterprise software procurement will increasingly prioritize agents that offer deep environmental integration, verifiable memory retention, and low management friction.

Redefining the Nature of Digital Workplace Collaboration

Teammates highlight an ongoing structural transformation across knowledge-work sectors. By equipping agents with persistent memory, role-based adaptability, and device-level deployment, Perplexity is establishing a blueprint for human-AI team configurations. The capability to delegate sustained responsibilities to digital entities frees human professionals from administrative friction, allowing teams to focus on strategy and decision-making.

At the same time, the deployment of always-on agents executing tasks on personal and corporate devices raises critical questions regarding workplace governance, data security, and task auditing. As organizations integrate persistent agents into shared workspaces, companies will need robust visibility into how memory is retained, how permissions are managed, and how autonomous actions on local devices are supervised.

Frequently Asked Questions

What are Perplexity Teammates and how do they function in a workspace?

Perplexity Teammates are always-on AI agents created specifically for workplace productivity. They are capable of adopting multiple roles across an organization's workspace, operating persistently rather than solely in response to individual prompts. Teammates accomplish complex assignments by drawing on existing projects, pre-configured skills, and accumulated memory from the Perplexity tools an organization already utilizes.

How does Perplexity Computer integrate with Teammates?

Perplexity Computer serves as the foundational AI system that enables the deployment of Teammates directly onto a user's device. Through this integration, Teammates are not limited to passive browser interactions; they can execute tasks within the local computing environment, bridging cloud-based AI reasoning with native desktop operations.

How does this launch change Perplexity's competitive position in the AI market?

The release of Teammates positions Perplexity in direct competition with frontier artificial intelligence developers like OpenAI and Anthropic in the enterprise agent market. By moving from search and synthesis into persistent, multi-role autonomous workflow execution, Perplexity is actively vying to become the core operating layer for enterprise work automation.

Related News

Gemini API Tool Retrieval Preview Gives Agents Access to Hundreds of Tools With Fewer Tokens
Product Launch

Gemini API Tool Retrieval Preview Gives Agents Access to Hundreds of Tools With Fewer Tokens

Google AI Blog announced that Tool Retrieval is now in preview in the Gemini API. The feature is presented as a way to give agents access to hundreds of tools while using fewer tokens. The announcement pairs those two points: the breadth of tool access and lower token use. It does not explain how Tool Retrieval works, specify how token use was measured, or quantify the reduction. It also does not provide an exact tool count beyond describing access to hundreds of tools. The stated release status is preview, and the announcement does not give a timeline for general availability.

LangSmith Tool Gateway Enters Public Beta for Secure, Governed AI Agent Tool Use
Product Launch

LangSmith Tool Gateway Enters Public Beta for Secure, Governed AI Agent Tool Use

LangChain announced that LangSmith Tool Gateway is in Public Beta. The gateway handles credentials, identity, and access for every agent tool call, with policies, approvals, and traces. The announcement frames secure, governed tool use around the individual call: its focus is on how agents access tools, rather than on what those tools or agents can do. The update identifies what the gateway covers but does not explain how credentials are managed, how access decisions are made, when approvals apply, or what information traces contain. The Public Beta is the only release status stated in the announcement.

Google’s Gemini Decision 1 Picks Answers Instead of Writing Them Across Five Input Types
Product Launch

Google’s Gemini Decision 1 Picks Answers Instead of Writing Them Across Five Input Types

Google’s Gemini Decision 1 is a model designed to pick answers rather than write them, according to Shittu Olumide’s report for KDnuggets. It reads text, images, video, audio, and PDFs in a single forward pass and supports up to 262,144 input tokens per request. The central distinction is between what the model reads and what it does with that input: it accepts several kinds of material, but its stated task is answer selection, not answer writing. The token figure describes the maximum input per request, not how much the model outputs. The report does not explain how answers are presented for selection or provide results showing how well the model performs.