
Teammates and Multiagent are now in the Perplexity app on iOS and Android, with voice for Teammates. About a third of the most demanding Computer work already starts on iPhone. Now control Teammates with voice. Available today for Max subscribers.

product launches + social @perplexity_ai | endlessly curious

Teammates and Multiagent are now in the Perplexity app on iOS and Android, with voice for Teammates. About a third of the most demanding Computer work already starts on iPhone. Now control Teammates with voice. Available today for Max subscribers.

We are live! Join us for an AMA with Perplexity's CTO, @denisyarats, and lead researcher on decision models, @antoine_chaffin @shashankgoyal95 will be chatting through Decider V1.3, Perplexity's new decision model that's free on OpenRouter right now, where search and AI are headed, and more. x.com/i/broadcasts/2jPqVvgrL…

Yesterday, I revealed that Perplexity crossed $1B ARR. We have strong margins and a healthy growth rate. We don’t talk about our finances very often because we’re focused on building. Follow our research: research.perplexity.ai

We’re rolling out Multiagent, a new effort level in Computer and teammates. A lead agent splits tasks across a team of agents using different frontier models. As they work, agents share findings with one another, with the lead agent combining everything into one answer.

I’ve been in this situation once in my life. Big combined swell and no beach to paddle into. And I was sure I was done for. Its not even the huge waves that are the most dangerous during a swell like this. The amount of water moving beneath the surface makes it extremely difficult to swim and stay afloat. Even really strong swimmers can easily drown in conditions like this. It wears you out quick and you have to do everything you can to stay calm and look for a way out. The only thing that saved me was that my leash didn’t break and was able to slowly paddle my way out before the next set. But I vividly remember thinking about my family and how crazy it was that I was going to die this way. People underestimate just how dangerous conditions like this are. Getting caught inside in the wrong conditions is deadly. This dude was extremely lucky. California lifeguards are elite and absolutely saved this guys life.
La Jolla, CA (NOV.15.2026)- One of the most extraordinary man versus ocean scenes ever captured. A solo lifeguard fights his way into the surf to save a stranded body border. There is no beach to get to. The video speaks for itself as the lifeguard and the body boarder are eventually saved by a lifeguard jet ski. Video 2x speed (full length in comments) 🎥 sailor_rea on TT

Announcing Perplexity Frontier Research. We are committing at least 10,000 GPUs to work on multi-agent orchestration and continual learning systems with the goal of building an automated researcher. We will publish evals, papers, blogs and open source our models and tooling. We’re actively hiring!
Today we're launching Perplexity Frontier Research. The goal of Perplexity Frontier Research is to create a system that can perform autonomous research. The work will focus on multi-agent orchestration, reinforcement learning, continual learning, systems, and pre-training. We will develop algorithms, models, and evaluations, released openly, and tested in Perplexity products. Our work will be supported by more than 10,000 GPUs. The Frontier Research team will regularly publish papers, models, data, and benchmarks. We're actively hiring researchers and engineers for Perplexity Frontier Research in San Francisco, Palo Alto, New York, and London. research.perplexity.ai

Today we're launching Perplexity Frontier Research. The goal of Perplexity Frontier Research is to create a system that can perform autonomous research. The work will focus on multi-agent orchestration, reinforcement learning, continual learning, systems, and pre-training. We will develop algorithms, models, and evaluations, released openly, and tested in Perplexity products. Our work will be supported by more than 10,000 GPUs. The Frontier Research team will regularly publish papers, models, data, and benchmarks. We're actively hiring researchers and engineers for Perplexity Frontier Research in San Francisco, Palo Alto, New York, and London. research.perplexity.ai


Perplexity Teammates is really growing on me, really dig that I can select model and effort level. This has been one of my biggest challenge with other ai assistants, I think it’s limiting if you’re using the same model and effort level for everything. For anyone following the benchmarking I’ve been doing with @VulcanBench you know how deep I’ve gone into really understanding what model and effort level you need based on things like task complexity. With ai personal assitants, I don’t need the same model/effort level to do a daily email summary as I do for more complex stock research or code review. I know in the future, the idea is that labs will determine what model/effort level you need, but right now it’s clear to me that most ai assistants just use the same model and effort level for everything…except for Perplexity. Personally I’d like to see Grok Bot, Muse, etc. offer something similar. This is pulling me back into the Perplexity camp after quite a hiatus on my part.

A week later, it’s 11x more agent hours per week from Teammate. Growing more than 50% week over week (was 7x last week). The number of employees getting an extra workday (8+ hours) and extra work week (40+ hours) from Teammate has more than tripled in a week since September. That’s roughly $80M to $180M a year in labor value against about $10M in tokens, so 9x to 19x depending on the job.
The median Perplexity employee now gets 7x more agent-hours per week from teammates than in September. And this has been consistently up every single week. 1 in 5 are already at 40+ agent-hours a week, which is basically a full extra workweek.

Teammates are now available in the Perplexity Mac app for Max subscribers. Teammates take on work for you. They work individually or with other teammates to handle QA, customer service, research, growth marketing, and more.

The browser agent in Computer and teammates now shows you what it’s working on. Watch live, take over, and securely sign in when needed, then hand control back to the agent.

We're hiring a Perplexity Physical Maker to work in our San Francisco office using equipment like the Bantam Tools NextDraw, Universal Robots UR5e, Makera Carvera, and Bambu Lab H2D to fabricate physical artifacts. You'll brainstorm ideas, order and set up machines in the office, connect them to Perplexity to orchestrate the creation of artifacts, record video of the results, and help your coworkers make their own creations. Our ideal candidate has deep experience using AI for physical creation, a love of iterative experimentation, and infectious enthusiasm for trying and sharing new tools. Contract role with potential to convert to fulltime, 40+ hours/week. To apply, use the Google form linked in the reply.

Persistent Agents are the next unlock of AI usage Our spend on AI is now 45% persistent agents Unfortunately for Anthropic, Claude Tags sucks OpenAI's Slack agent is even worse We primarily use Perplexity Computer, which uses a lot of different models Perplexity's Harness mogs

We are publishing new research on our findings from red-teaming SPACE, our VM sandbox for Computer agents. We gave 12 models two tasks: escape the VM to take a credential from the host, and bypass SPACE's network policy to reach a blocked URL. We found that the standard Capture the Flag setup used in red-teaming can steer agents away from the security boundary being tested, and that security-specialized models don't guarantee better results. perplexity.ai/hub/blog/escap…

Unlocking most enterprise workflows is now a context and connectors problem, not a model one. The two key challenges I see most commonly are 1) getting enterprises to feel secure bringing in that context and 2) not having agents waste valuable tokens finding the right data.
The bottleneck of automation for 99% of enterprises is no longer smarter LLMs... The current LLMs are capable enough New models coming out now mostly just mean defaulting to higher costs. We force default earlier models unless there is a strong reason they need the latest.