Aravind Srinivas(@AravSrinivas)

Perplexity Computer makes you money.

Aravind Srinivas (@AravSrinivas) — Fri, 11 Sep 2026 17:07:44 GMT

Perplexity Computer makes you money.

Coinbase Developer Platform🛡️@CoinbaseDev

Where are agentic traders coming from?

Here's the breakdown of where agentic trading volume came from last week:

Perplexity - 26.8%
Claude - 19.3%
Grok - 15.3%
ChatGPT - 1.1%
Claude Code - 0.7%
CLI/Other - 36.8%

🔗 View Quoted Tweet

💬34🔄23❤️417👀46172📊61

⚡ Powered by xgo.ing

💚

Aravind Srinivas (@AravSrinivas) — Thu, 10 Sep 2026 22:21:27 GMT

💚

NVIDIA AI@NVIDIAAI

Another leaderboard win for Nemotron 🏆

Nemotron 3 Embed 8B ranks #1 for combined nDCG@10 on the Q2D-Web benchmark, tested across 190M web documents and nearly 70K agent-reformulated queries in 10 languages.

Shoutout to @perplexity_ai for putting it together.

🔗 View Quoted Tweet

💬8🔄7❤️210👀33375📊20

⚡ Powered by xgo.ing

🦀🦀🦀

Aravind Srinivas (@AravSrinivas) — Thu, 10 Sep 2026 19:55:52 GMT

🦀🦀🦀

Perplexity Developers@perplexitydevs

Perplexity has joined the Rust Foundation.

We believe in supporting the people who build reliable open-source software. By joining the Rust Foundation, our goal is to improve how people and agents build with Rust.

🔗 View Quoted Tweet

💬28🔄24❤️819👀50417📊70

⚡ Powered by xgo.ing

one aspires to be able to ship something like this in their lifetime. art and craftsmanship.

Aravind Srinivas (@AravSrinivas) — Thu, 10 Sep 2026 07:38:25 GMT

one aspires to be able to ship something like this in their lifetime. art and craftsmanship.

Your browser does not support the video tag.

🔗 View on Twitter

💬117🔄188❤️5628👀327258📊584

⚡ Powered by xgo.ing

the perfect pocket computer does not exi…

Aravind Srinivas (@AravSrinivas) — Thu, 10 Sep 2026 07:08:31 GMT

the perfect pocket computer does not exi…

💬51🔄20❤️1157👀130792📊132

⚡ Powered by xgo.ing

Wow!

Aravind Srinivas (@AravSrinivas) — Thu, 10 Sep 2026 06:22:29 GMT

Wow!

DeepSeek@deepseek_ai

🚀 Introducing DeepSeek-V4.1-Flash: smarter, faster, more efficient.

🔹 Introducing the smallest model in our new architecture family, with native visual understanding.
🔹 Designed for greater capability, faster inference, higher throughput, and scaling to larger models.

1/6

🔗 View Quoted Tweet

💬34🔄63❤️1462👀257897📊134

⚡ Powered by xgo.ing

😍

Aravind Srinivas (@AravSrinivas) — Thu, 10 Sep 2026 03:14:18 GMT

😍

Greg Joswiak@gregjoz

It’s THIN.
It’s VERSATILE
It’s DURABLE.
It fits in your pocket.

And it FOLDS.

Meet iPhone Duo.

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬12🔄10❤️741👀111765📊54

⚡ Powered by xgo.ing

Web app development usage (particularly those that require pulling information from several differen...

Aravind Srinivas (@AravSrinivas) — Wed, 09 Sep 2026 16:09:29 GMT

Web app development usage (particularly those that require pulling information from several different places) is expanding quickly on Perplexity Computer. We're now supporting mobile and desktop previews for the websites created in the thread.

Computer@AskPerplexity

Websites created with Computer now have desktop and mobile previews.

Use the artifact side pane to switch between views, refresh the preview, edit the site, or leave a comment.

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬22🔄15❤️195👀60198📊29

⚡ Powered by xgo.ing

Perplexity Search inside Hermes. Perplexity's index currently includes 450B+ high-quality URLs. Rapi...

Aravind Srinivas (@AravSrinivas) — Wed, 09 Sep 2026 02:24:01 GMT

Perplexity Search inside Hermes. Perplexity's index currently includes 450B+ high-quality URLs. Rapidly advancing to a trillion by EOY with high-quality snippets.

Perplexity Developers@perplexitydevs

Perplexity Search API is now available in Hermes Agent.

Search API gives Hermes access to an index of more than 400 billion URLs. It returns real-time results and ranks snippets by relevance.

pplx.ai/hermes

🔗 View Quoted Tweet

💬36🔄50❤️589👀92782📊80

⚡ Powered by xgo.ing

I have admired @Teknium and @NousResearch for relentless product execution and taste. Really glad to...

Aravind Srinivas (@AravSrinivas) — Wed, 09 Sep 2026 02:18:39 GMT

I have admired @Teknium and @NousResearch for relentless product execution and taste. Really glad to have Perplexity Search inside the Hermes Agent harness!

Nous Research@NousResearch

You can now use the Perplexity Search API in Hermes Agent

🔗 View Quoted Tweet

💬25🔄26❤️728👀80973📊65

⚡ Powered by xgo.ing

Monumental accomplishment!

Aravind Srinivas (@AravSrinivas) — Tue, 08 Sep 2026 19:32:35 GMT

Monumental accomplishment!

OpenAI@OpenAI

We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.

The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.

The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.

🔗 View Quoted Tweet

💬31🔄38❤️1202👀84660📊100

⚡ Powered by xgo.ing

We have started serving a lot of our inference on NVLink Blackwells. Vera Rubins next!

Aravind Srinivas (@AravSrinivas) — Tue, 08 Sep 2026 17:10:38 GMT

We have started serving a lot of our inference on NVLink Blackwells. Vera Rubins next!

NVIDIA AI Infrastructure@NVIDIAAIInfra

What does it take to deliver high-quality results in seconds at massive scale?

@perplexity_ai serves 50M+ queries daily across 15 models, including NVIDIA Nemotron 3 Ultra. That takes high-performance infrastructure and smarter routing without compromising quality, speed, or cost.

That’s why Perplexity has built on NVIDIA generation after generation. With NVIDIA Blackwell, the larger NVLink domain enables models to scale across multiple nodes without sacrificing latency.

Learn more: nvda.ws/3SSC4D6

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬24🔄22❤️325👀44559📊39

⚡ Powered by xgo.ing

Detecting malicious intent of agents and performing forensics is going to be crucial considering wha...

Aravind Srinivas (@AravSrinivas) — Sat, 05 Sep 2026 04:07:47 GMT

Detecting malicious intent of agents and performing forensics is going to be crucial considering what’s recently happened with rogue agents escaping sandboxes and attacking third-party sites. Numbat is Perplexity’s open-source tool that can help defenders. holisticinfosec.io/post/numbat/

💬65🔄24❤️292👀39057📊84

⚡ Powered by xgo.ing

Perplexity Pro and Max subscribers get to use both Fable and Astra on Computer mode. Enjoy!

Aravind Srinivas (@AravSrinivas) — Sat, 05 Sep 2026 03:41:11 GMT

Perplexity Pro and Max subscribers get to use both Fable and Astra on Computer mode. Enjoy!

Perplexity@perplexity_ai

GPT-6 Astra is now available in Perplexity Computer for Pro and Max subscribers.

🔗 View Quoted Tweet

💬69🔄68❤️1853👀1267460📊169

⚡ Powered by xgo.ing

Join us if you want to work on hard inference and infrastructure engineering problems!

Aravind Srinivas (@AravSrinivas) — Sat, 05 Sep 2026 02:38:16 GMT

Join us if you want to work on hard inference and infrastructure engineering problems!

Denis Yarats@denisyarats

check out our new blog post on how we serve embeddings and rerankers for our SOTA pplx-embed models over an exabyte-scale search index.

embedding models have a similar serving profile to LLMs: batch indexing is compute-bound prefill, online serving is memory-bound decode. this lets us reuse the same optimized kernels we built for large LLMs and get great efficiency for free. on top of that, we optimized the runtime: whole-model CUDA graphs captured lazily as the engine serves, and a LazyTensor in Rust that overlaps CPU scheduling with GPU execution.

result: up to 3x lower p50 and 4.8x lower p99 latency than vLLM on BGE-M3 at 128 tokens, single H200!

if you are interested in working on problems like this, DM me or apply at perplexity.ai/hub/careers

🔗 View Quoted Tweet

💬30🔄8❤️419👀59533📊52

⚡ Powered by xgo.ing

A deep dive into how Perplexity serves search results at scale: embeddings for ranking, GPU-based mo...

Aravind Srinivas (@AravSrinivas) — Fri, 04 Sep 2026 21:33:32 GMT

A deep dive into how Perplexity serves search results at scale: embeddings for ranking, GPU-based model inference, request batching, running inference servers, and handling latency/throughput trade-offs.

Perplexity@perplexity_ai

Every answer in Perplexity starts with embedding and ranking models picking the most relevant results for the query.

Today we published research on how we built SoTA serving infrastructure behind those models.

Read the research: perplexity.ai/hub/blog/fast-…

🔗 View Quoted Tweet

💬35🔄29❤️619👀79901📊92

⚡ Powered by xgo.ing

Perplexity will be at RustConf to present how we built the sandboxes powering all of Perplexity Comp...

Aravind Srinivas (@AravSrinivas) — Fri, 04 Sep 2026 19:52:03 GMT

Perplexity will be at RustConf to present how we built the sandboxes powering all of Perplexity Computer.

Perplexity Developers@perplexitydevs

Perplexity will be at RustConf 2026 next week.

@zbraniecki will present “Inside SPACE, the Rust Sandbox Platform Behind Perplexity Computer” on September 10. Then join us for happy hour from 5:30–7:30 PM.

Register at:
pplx.ai/rc-happy-hour

🔗 View Quoted Tweet

💬14🔄12❤️293👀40664📊30

⚡ Powered by xgo.ing

the end of markdown files. bitter lesson always wins.

Aravind Srinivas (@AravSrinivas) — Fri, 04 Sep 2026 19:03:56 GMT

the end of markdown files. bitter lesson always wins.

💬269🔄68❤️4277👀913400📊665

⚡ Powered by xgo.ing

Astra is also awesome at computer use. We will be bringing it to Comet, as well as our cloud browser...

Aravind Srinivas (@AravSrinivas) — Thu, 03 Sep 2026 21:30:34 GMT

Astra is also awesome at computer use. We will be bringing it to Comet, as well as our cloud browser sandbox that powers all Computer usage. Evals coming soon!

💬45🔄46❤️1560👀58091📊158

⚡ Powered by xgo.ing

Congrats to @OpenAI on building the industry's frontier model: GPT-6 Astra. It's far ahead of every ...

Aravind Srinivas (@AravSrinivas) — Thu, 03 Sep 2026 21:13:22 GMT

Congrats to @OpenAI on building the industry's frontier model: GPT-6 Astra. It's far ahead of every other model on wide and deep research tasks, while also being more cost-effective. We'll be bringing this model up on Perplexity Computer for all Pro and Max users soon!

Perplexity@perplexity_ai

We evaluated GPT-6 Astra on WANDR. It scored 0.682 at $11.98 per task, the highest score of any model we tested.

GPT-6-Astra scored 13.5% higher than Fable 5.1 at 6.1% lower cost, and 27.0% higher than Opus 5 at 3.3% higher cost.

🔗 View Quoted Tweet

💬62🔄135❤️3138👀220922📊264

⚡ Powered by xgo.ing

This is cool. We need more projects of this nature to address the power and memory/compute bottlenec...

Aravind Srinivas (@AravSrinivas) — Thu, 03 Sep 2026 18:58:23 GMT

This is cool. We need more projects of this nature to address the power and memory/compute bottlenecks that stop us from scaling the adoption of agents.

NVIDIA RTX Spark@NVIDIARTXSpark

Your devices are stronger together. 🖥️🤝🖥️

Just announced at IFA, NVIDIA PAIR automatically links systems across your local network and sends inference requests wherever there’s available capacity, helping agents run more efficiently.

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬28🔄30❤️552👀70744📊72

⚡ Powered by xgo.ing

Portable Computer (fully local runtime of Perplexity Computer) is now compatible with RTX (Linux). W...

Aravind Srinivas (@AravSrinivas) — Thu, 03 Sep 2026 16:32:26 GMT

Portable Computer (fully local runtime of Perplexity Computer) is now compatible with RTX (Linux). Windows coming next!

Perplexity@perplexity_ai

Portable Computer is now available on Linux for @NVIDIA RTX GPUs with 24GB of VRAM or higher.

🔗 View Quoted Tweet

💬25🔄22❤️339👀42425📊41

⚡ Powered by xgo.ing

A repository of open models and tools to train and serve them in an accessible manner is absolutely ...

Aravind Srinivas (@AravSrinivas) — Thu, 03 Sep 2026 14:24:48 GMT

A repository of open models and tools to train and serve them in an accessible manner is absolutely necessary for AI to remain accessible and useful to the public. Glad that NVIDIA is coming forth to do that for the open source community by supporting HuggingFace.

Jensen Huang@JensenHuang

Exciting day for NVIDIA and @huggingface.

Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty. They allow every developer, startup, university, industry and country to build with, customize and benefit from AI.

Thank you @ClementDelangue for coming to me.

NVIDIA is going to be a great home for Hugging Face, its community and the future of open models. 🤗
blogs.nvidia.com/blog/nvidia-to…e

🔗 View Quoted Tweet

💬33🔄26❤️533👀53825📊58

⚡ Powered by xgo.ing

open-source RL-as-a-service https://t.co/UdKabiSfJP

Aravind Srinivas (@AravSrinivas) — Thu, 03 Sep 2026 03:33:03 GMT

open-source RL-as-a-service github.com/radixark/miles

💬29🔄74❤️867👀79255📊217

⚡ Powered by xgo.ing

https://t.co/bY7iZGmcsA

Aravind Srinivas (@AravSrinivas) — Wed, 02 Sep 2026 21:38:09 GMT

github.com/perplexityai/p…

💬5🔄3❤️52👀9402📊11

⚡ Powered by xgo.ing

We're open-sourcing Lily, Perplexity's local inference engine for serving models locally on Apple Si...

Aravind Srinivas (@AravSrinivas) — Wed, 02 Sep 2026 21:37:37 GMT

We're open-sourcing Lily, Perplexity's local inference engine for serving models locally on Apple Silicon. This powers Perplexity's newly introduced hybrid compute feature for the Mac app.

Perplexity@perplexity_ai

Today we’re open-sourcing Lily, the local inference engine we built for hybrid compute in Perplexity Computer.

Lily is specialized for Qwen3.6-35B-A3B on Apple silicon, built so on-device compute doesn’t bottleneck Computer tasks.

Read more: perplexity.ai/hub/blog/optim…

🔗 View Quoted Tweet

💬74🔄134❤️1856👀137105📊249

⚡ Powered by xgo.ing

Live demo of Perplexity Computer running fully local on DGX Spark

Aravind Srinivas (@AravSrinivas) — Wed, 02 Sep 2026 19:35:27 GMT

Live demo of Perplexity Computer running fully local on DGX Spark

NVIDIA AI@NVIDIAAI

DGX Spark Live: Perplexity Portable Computer Goes Local x.com/i/broadcasts/1…

🔗 View Quoted Tweet

💬20🔄18❤️405👀51207📊49

⚡ Powered by xgo.ing

Private, Personal, Powerful AI.

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 23:18:42 GMT

Private, Personal, Powerful AI.

Kyle Polley@kpolley

Best-in-class PII Detection...running entirely on your device 🤯 
perplexity.ai/hub/blog/pii-t…A

🔗 View Quoted Tweet

💬18🔄23❤️273👀44620📊38

⚡ Powered by xgo.ing

This is an important observation. Agents will get smart enough to spin new on-demand GPU nodes to go...

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 20:46:45 GMT

This is an important observation. Agents will get smart enough to spin new on-demand GPU nodes to go train themselves with minimal oversight. Inserting sufficient guardrails and friction here is necessary.

💬54🔄39❤️889👀94788📊131

⚡ Powered by xgo.ing

Live demo of Portable Computer (aka a fully local runtime of Perplexity Computer) on DGX Spark

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 20:29:13 GMT

Live demo of Portable Computer (aka a fully local runtime of Perplexity Computer) on DGX Spark

NVIDIA RTX Spark@NVIDIARTXSpark

Want to see @perplexity_ai Portable Computer running locally on NVIDIA DGX Spark? 👀

Tune in tomorrow, 9/2 at 12:30 p.m. PDT, to see live demos showcasing private file analysis, connected tools and more.

🔗 View Quoted Tweet

💬14🔄4❤️178👀37219📊21

⚡ Powered by xgo.ing

Agentic trading, market analysis, and monitoring on Perplexity Computer with Coinbase

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 19:55:40 GMT

Agentic trading, market analysis, and monitoring on Perplexity Computer with Coinbase

Coinbase Developer Platform🛡️@CoinbaseDev

Coinbase for Agents is now live on Perplexity Computer.

Let your Computer trade, analyze markets, and execute workflows. One of 20+ financial connectors.

No APIs. No MCP setup. Just click connect.

@perplexity_ai Computer works for you. Now it works with Coinbase.

🔗 View Quoted Tweet

💬20🔄17❤️286👀43294📊39

⚡ Powered by xgo.ing

Fable is the frontier model by a good margin right now. Enjoy it on Perplexity Computer as an orches...

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 19:14:29 GMT

Fable is the frontier model by a good margin right now. Enjoy it on Perplexity Computer as an orchestrator for high-stakes tasks. Computer still uses GPT 5.6 (Terra) models as cost-efficient subagents while taking advantage of Fable's frontier planning capabilities.

💬55🔄30❤️530👀39726📊92

⚡ Powered by xgo.ing

And pointers to get started with hybrid compute on Perplexity Mac App: https://t.co/wDwvO57luO

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 15:04:37 GMT

And pointers to get started with hybrid compute on Perplexity Mac App: perplexity.ai/hub/products/h…

💬4🔄3❤️99👀12923📊18

⚡ Powered by xgo.ing

Research blog on privacy protecting local runtime: https://t.co/3wsCZByISY

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 15:04:36 GMT

Research blog on privacy protecting local runtime: perplexity.ai/hub/blog/pii-t…

💬1🔄2❤️101👀14310📊14

⚡ Powered by xgo.ing

We're introducing hybrid compute for all users of the Perplexity Mac app. This will allow Computer...

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 15:04:34 GMT

We're introducing hybrid compute for all users of the Perplexity Mac app. 

This will allow Computer to orchestrate local models that can run locally on Mac, particularly for agent steps involving sensitive and private files (eg your bloodwork, tax returns, litigation, etc).

Your browser does not support the video tag.

🔗 View on Twitter

💬227🔄280❤️3936👀496844📊680

⚡ Powered by xgo.ing

Most AI agent apps don't differentiate between native and web-based runtimes when using cloud-based ...

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 15:04:34 GMT

Most AI agent apps don't differentiate between native and web-based runtimes when using cloud-based agents. Mac offers a big opportunity to move token consumption to Apple Silicon (with no price paid for tokens consumed locally) and protect user privacy. Hybrid compute combines the best of cloud-based frontier models and a privacy-protecting local runtime.

💬6🔄4❤️198👀26738📊21

⚡ Powered by xgo.ing

excited 4 tmrw 

Aravind Srinivas (@AravSrinivas) — Tue, 01 Sep 2026 04:43:22 GMT

excited 4 tmrw  

💬176🔄74❤️4254👀649822📊484

⚡ Powered by xgo.ing

Congrats Tim on an incredible journey at Apple. Few people have ever shipped at that scale without l...

Aravind Srinivas (@AravSrinivas) — Mon, 31 Aug 2026 16:42:46 GMT

Congrats Tim on an incredible journey at Apple. Few people have ever shipped at that scale without letting the bar for excellence slip. You've raised the standard for all of us to aspire to; thanks for everything.

Tim Cook@tim_cook

Sending lots of love to the Apple community on my last day as CEO. My title changes tomorrow, but the love I have for the Apple community never will. Thank you for being a constant source of inspiration. My gratitude is endless, and I’m excited for the next chapter!

🔗 View Quoted Tweet

💬67🔄176❤️6066👀319598📊426

⚡ Powered by xgo.ing

perplexity is the best at search at any level of compute, agnostic of how much compute the competito...

Aravind Srinivas (@AravSrinivas) — Fri, 28 Aug 2026 21:26:49 GMT

perplexity is the best at search at any level of compute, agnostic of how much compute the competitors use.

Artificial Analysis@ArtificialAnlys

Perplexity Search debuts on the Artificial Analysis Search Index, with all three context size variants taking top positions on the leaderboard

The @perplexity_ai Search API comes with three context settings (low, medium, and high) that control how much extracted content each search result carries. We tested all three variants using our standardized methodology: the same model (GPT-5.6 Luna at medium reasoning), running inside Stirrup, our open-source agent harness, with tools for searching and fetching pages from the web. Only the provider behind the search tool changes.

Key results:

➤ Perplexity Search (medium) scores 80 on the Artificial Analysis Search Index, ahead of the previous leaders, Parallel (advanced) and Brave Search (LLM context), at 75. The high and low variants score 79 and 77 respectively. Its lead is concentrated in BrowseComp results, with AA-Omniscience and DeepSearchQA scoring comparably to other leading providers

➤ Efficient search payloads: smaller overall search results mean the model reads less per task, so Perplexity has the lowest model inference cost per task of providers we’ve tested so far, ranging from $0.028 to $0.034 across the three variants vs $0.036 for the next lowest provider

➤ Total cost per task is ~$0.091 for the medium and high context variants, at mid-pack latency. For comparison, Parallel (advanced) costs $0.084 per task and Brave (LLM context) costs $0.13 per task

🔗 View Quoted Tweet

💬37🔄26❤️457👀223848📊57

⚡ Powered by xgo.ing

Decagon is powering customer support for some of the biggest brands in the industry: Delta Airlines,...

Aravind Srinivas (@AravSrinivas) — Fri, 28 Aug 2026 17:25:21 GMT

Decagon is powering customer support for some of the biggest brands in the industry: Delta Airlines, Ticketmaster, Deutsche Telekom, American Airlines. A lot of support related questions require accurate online information and we’re happy to power that with Perplexity’s search.

Decagon@DecagonAI

We’re partnering with @Perplexity_AI to bring live web search to Decagon agents.

Customer questions often depend on information that changes by the hour.

Agents can now search the live web mid conversation, pull in current information, and respond with cited sources.

🔗 View Quoted Tweet

💬17🔄18❤️200👀43956📊26

⚡ Powered by xgo.ing

Agentic Trading on Perplexity Computer with @public connector

Aravind Srinivas (@AravSrinivas) — Thu, 27 Aug 2026 15:57:10 GMT

Agentic Trading on Perplexity Computer with @public connector

Public@public

Public + Perplexity are officially connected. 

Public members can now connect their brokerage accounts to Perplexity to research and trade stocks, options, crypto, bonds, and Treasuries—all in one AI-powered experience. x.com/i/article/2092…

🔗 View Quoted Tweet

💬25🔄21❤️239👀42123📊41

⚡ Powered by xgo.ing

The Agentic Cloud

Aravind Srinivas (@AravSrinivas) — Thu, 27 Aug 2026 14:48:25 GMT

The Agentic Cloud

Perplexity Developers@perplexitydevs

Connectors are now available in Agent API.

Connect your agents to GitHub, Slack, Google Drive, and Datadog without passing a server URL or token on every request.

API Group admins can connect each service once for all API Group members.

pplx.ai/api-console

🔗 View Quoted Tweet

💬19🔄17❤️179👀223619📊27

⚡ Powered by xgo.ing

Hardware like DGX Spark, with a local agentic computer, will become the gateway for consuming fronti...

Aravind Srinivas (@AravSrinivas) — Wed, 26 Aug 2026 19:29:40 GMT

Hardware like DGX Spark, with a local agentic computer, will become the gateway for consuming frontier tokens.

Denis Yarats@denisyarats

at perplexity we are really excited about multi-agent collaboration! we recently explored a simple instantiation of this with advisor escalation from a local model to a remote frontier model, where we showed it can significantly boost the local model's performance.

an alternative way i like to think about this: the local model becomes a **gateway** to remote models. its main job is to preprocess all the raw tokens, package them into fewer, more information-dense tokens, and send only those to the server. this preprocessing has several advantages: it reduces cost and latency, enables personalization, and allows privacy controls. already such approaches can save 50% of tokens, but imagine a future where 90%+ of tokens are processed locally and only 10% are sent to the server for the last-mile hardest reasoning. that is a 10x cost reduction and a much more pleasant UX due to low latency.

training such a multi-agent system is not an easy problem and still requires a lot of research in multi-agent RL, but we are making good progress and actively hiring for this. if this sounds interesting please DM me

🔗 View Quoted Tweet

💬29🔄17❤️209👀38116📊35

⚡ Powered by xgo.ing

We are doubling down on making Computer more useful for financial researchers and analysts. Comput...

Aravind Srinivas (@AravSrinivas) — Wed, 26 Aug 2026 18:20:31 GMT

We are doubling down on making Computer more useful for financial researchers and analysts. 

Computer now connects to new licensed data sources, including Dun & Bradstreet, Guidepoint, IBISWorld, and 20+ others. An analyst can ask Computer a question and get an answer built from the firm's own licensed sources without APIs or separate logins to work through. 

Every figure traces to the source record it came from. Teams doing high-stakes work would know exactly which data set produced which number.

To connect, select a provider in Settings → Connectors and sign in with an existing license. Available now to all Perplexity users.

perplexity.ai/hub/blog/compu…

💬38🔄51❤️673👀349262📊139

⚡ Powered by xgo.ing

Perplexity Computer for Max users comes with a background Dream agent that continually ingests conte...

Aravind Srinivas (@AravSrinivas) — Wed, 26 Aug 2026 15:35:45 GMT

Perplexity Computer for Max users comes with a background Dream agent that continually ingests context from files and connected apps and builds multi-hop context graphs in a perpetual compounding loop. Results show a significant jump in correctness, recall and token efficiency.

Perplexity@perplexity_ai

Brain is our self-improving memory system for Perplexity Computer. It compiles sessions, files, and sources into a structured knowledge wiki.

New evals build on our initial results, improving correctness by 9.3 points, currentness by 8.0, and recall by 8.9 with 15% fewer tokens.

🔗 View Quoted Tweet

💬41🔄26❤️422👀67932📊75

⚡ Powered by xgo.ing

A background process that continually ingests context from every single connector (or app), performs...

Aravind Srinivas (@AravSrinivas) — Wed, 26 Aug 2026 01:47:38 GMT

A background process that continually ingests context from every single connector (or app), performs multi-hop reasoning in a perpetual inference loop, and runs on your hardware. That is the future.

💬79🔄36❤️755👀63466📊149

⚡ Powered by xgo.ing

Big thanks to @nvidia for working together with us and supporting this research on DGX Spark as well...

Aravind Srinivas (@AravSrinivas) — Tue, 25 Aug 2026 20:08:05 GMT

Big thanks to @nvidia for working together with us and supporting this research on DGX Spark as well as enabling an open ecosystem around cost-effective open-weight models, inference frameworks, and hardware with unified memory.

Perplexity@perplexity_ai

New research: Portable Computer is a local-first agent for private and cost-effective work.

With an on-device 27B model, our harness scores 82.6% on real knowledge work, beating open-source harnesses Pi and Hermes. Our post-trained PPLX 27B reaches 85.4%.

🔗 View Quoted Tweet

💬12🔄24❤️315👀37261📊32

⚡ Powered by xgo.ing

On showing an early demo of Portable Computer on DGX Spark to Jensen, he was kind to gift us a DGX S...

Aravind Srinivas (@AravSrinivas) — Tue, 25 Aug 2026 15:57:11 GMT

On showing an early demo of Portable Computer on DGX Spark to Jensen, he was kind to gift us a DGX Station, a beast of a local computer that can serve even frontier models like GLM 5.3. Unmetered frontier intelligence running on your own local hardware coming soon!

💬70🔄69❤️1548👀88374📊195

⚡ Powered by xgo.ing

Portable Computer on Spark blog: https://t.co/Ze310Yl9hq

Aravind Srinivas (@AravSrinivas) — Tue, 25 Aug 2026 15:27:27 GMT

Portable Computer on Spark blog: 
perplexity.ai/hub/products/p…

💬2🔄0❤️17👀4515📊3

⚡ Powered by xgo.ing

Research Blog: https://t.co/UVow5cJgcO

Aravind Srinivas (@AravSrinivas) — Tue, 25 Aug 2026 15:24:46 GMT

Research Blog: perplexity.ai/hub/blog/a-loc…

💬1🔄2❤️18👀4797📊3