Perplexity
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 24
- Perplexity Decider v1.2 and Decider Flash v1.2 arrive on OpenRouter
OpenRouter said Perplexity's open-weights decision models, Decider v1.2 and Decider Flash v1.2, are now available on its platform. The models accept text or images and return typed answers with probabilities. Decider v1.2 is a new 27B checkpoint succeeding v1.1, while Decider Flash v1.2 is a 4B model that OpenRouter describes as faster and half the cost.
OpenRouter (X) · 🔥 8 - Oxylabs, Decodo and three more scraping brands unite under Source
Five proxy and web-scraping brands — Oxylabs, Decodo, Webshare, ScrapingBee and Honeygain — now formally sit under a single Vilnius-based umbrella called Source, after years of being sold and reviewed as rivals; three of them appear in TechRadar's own best-proxy picks. Source says demand across its companies rose fourfold in a year as AI agents, next-generation search tools and frontier AI products pull live data from the web, that its systems process 1.5 million requests per second, and that it is expanding into web search APIs built for AI agents. The group says it has no customers or products of its own, handling patents, procurement, hiring and compliance while each brand keeps its own team, product strategy and customer base; the report also notes that Oxylabs is a defendant in Reddit's scraping lawsuit.
TechRadar · 🔥 7 - UK SMEs' AI spending hits new high as Anthropic edges out OpenAI, study says
A study by small-business credit card provider Capital on Tap says UK SME adoption of AI tools rose from 1.1% in 2023 to 12.8% in 2026, with roughly one in eight small businesses paying for AI and average quarterly spending reaching £75.60 per SME. The research puts Anthropic's Claude and related tools at 54.1% of that AI spending and OpenAI/ChatGPT second at 27.3%, while more than 28% of SMEs use two or more platforms. It also claims publishing leads sector adoption at 26%, ahead of radio and television and consulting (21.6% each) and telecoms (20.3%), and advises businesses to cut duplicate subscriptions and check outputs.
TechRadar · 🔥 5 - Amazon blocks Grok Bot shopping; Musk floats an Amazon rival
Amazon has blocked Elon Musk's Grok Bot agent from placing orders on its site, telling users that unauthorized AI agents violate the terms they agreed to, according to several Chinese tech reports. Musk responded by suggesting users shop elsewhere for cheaper goods and wrote on X that "maybe we should create a company that competes with Amazon," though no product plan has been announced. The reports add that Amazon previously restricted Perplexity's Comet and Meta's Muse, lists programs such as ChatGPT-User, Claude-User and GoogleAgent-Shopping in robots.txt, and runs its own Rufus, Alexa for Shopping and Buy for Me, while Shopify has opened its catalog to more agents through Shop MCP.
36氪 人工智能 · 🔥 27 - Perplexity open sources SPACE, its AI agent sandbox
Perplexity is open sourcing SPACE, the AI agent sandbox it has been using to combat rogue AI agents. The company told Semafor that partners including CoreWeave, Akamai and Crusoe have agreed to offer SPACE to customers. The reports did not specify licensing terms, a release timeline, or technical details of the sandbox.
Techmeme · 🔥 4 - Nvidia's RTX Spark laptops launch with up to 128GB unified memory for local AI agents
Nvidia's RTX Spark laptops are now available from ASUS, Dell, HP, Lenovo, Microsoft and MSI. The platform packs a Grace CPU and Blackwell RTX GPU into one superchip (co-designed with MediaTek), with up to 128GB of unified memory and up to 1 petaflop of FP4 AI compute, which Nvidia says lets the machines run large language models and AI agents locally. A new Windows first-run experience installs agents from the Store in about two clicks, with Hermes, Perplexity and OpenClaw officially supported at launch. Nvidia also released the open-weight Nemotron 3.5 Bolt (74B parameters, 9B active, fitting in 24GB of memory); The Next Web cites an Artificial Analysis Intelligence Index score of 21, against 58 for the top model, Claude Opus 5.5. Companion tools include the Switchyard local/cloud model router, PAIR for serving models across a home network, and the OpenShell runtime plus Microsoft Execution Containers for agent controls. Lenovo's YOGA Pro 15 RTX Spark also went on sale in China starting at 29,999 yuan. The Verge notes Copilot+ features such as Recall are not available on RTX Spark at launch and are expected for Windows Insiders later this year.
The Verge AI · 🔥 33 - Getty Images files prepackaged Chapter 11 plan to go private
Getty Images said on October 12 that it has agreed a comprehensive restructuring with major financial stakeholders, using a prepackaged Chapter 11 plan to go private by the end of 2026. The package includes up to $533 million in new funding from the Getty family and Koch Equity Development, write-offs of $270 million in senior unsecured notes, and lower rates and extended maturities on other debt, while KED would provide up to $143 million more in restructuring financing; public shareholders' equity would be wiped out. The restructuring follows the July 7 termination of its roughly $3.7 billion merger with Shutterstock after UK regulators demanded a full editorial divestiture; 2025 revenue hit a record $981.3 million, up 4.5%, but net loss exceeded $200 million, and second-quarter 2026 revenue fell 2.5% to $229.1 million with a loss of $85.8 million. Generative AI has also hurt its licensing business: Getty sued Stability AI over training data, launched Generative AI by Getty Images with Nvidia, and signed multi-year content deals with Perplexity and OpenAI.
AIbase AI新闻 · 🔥 0 - Cohere pokes fun at Perplexity in a pun-filled X post
Cohere's official X account posted a joking message that plays on its own name and Perplexity's, saying roughly "Perplexity this, Cohere that, what if you said something coherent so she wasn't so perplexed." The post carries no product, model, or feature information, and no release date or version is mentioned.
cohere (X) · 🔥 0 - ChatPlayground AI lifetime deal bundles 20+ models
StackSocial is promoting a lifetime subscription to ChatPlayground AI, advertised at $60 for access to more than 20 AI models against a listed original price of $619, a claimed 90% discount. According to Gizmodo, the tool queries the models a user enables at the same time — including ChatGPT, Gemini, DeepSeek, Llama and Perplexity — with the option to turn individual models on or off, plus a prompt-improvement feature and customer support. The report says the plan includes unlimited messages per month under a fair-use policy, and that the service swaps in equivalent or updated model versions over time; the pricing and discount figures are the seller's claims, and the article notes all models can make mistakes that users should verify.
Gizmodo · 🔥 0 - Floqer pitches AI sales agents at Disrupt, plans to open-source database
Floqer is an autonomous customer knowledge platform that gives business-to-business sales and marketing teams an AI-powered database designed to continuously collect, organize and update information about their customers. The startup pitched on the Startup Battlefield stage at TechCrunch Disrupt 2026 and said it plans to open-source the technology behind its autonomous customer knowledge database, targeting early 2027 pending security reviews. Floqer says it has surpassed $1.2 million in annual recurring revenue, that overall revenue grew roughly tenfold over the past year, and that it has eight employees and more than 1,000 customers including Perplexity, Wise and AngelList; it also claims customers have cut prospecting time by as much as 70%.
TechCrunch AI · 🔥 0 - last30days agent skill adds native Codex plugin, new sources
The open-source AI agent skill last30days has updated to v3.11.1. According to the project, it has merged 175 pull requests since the v3.3 announcement in May — 122 of them from 52 community contributors across 15 releases. The skill lets an agent search Reddit, X, YouTube, Hacker News, Polymarket, GitHub and other platforms in parallel, rank results by upvotes, likes and money at stake, and synthesize one brief. This update adds a first-class OpenAI Codex plugin with renderer-aware citations, free keyless sources including arXiv, Techmeme and Digg, a rebuilt X search pipeline, and a doctor command that checks each source's health.
GitHub Trending · Python · 🔥 0 - Shopify opens its catalog to any AI agent for shopping and checkout
Shopify announced Tuesday that shoppers can configure their AI agents to browse and purchase products on their behalf, giving more agents access to more than a billion listings in its catalog, which includes items both on and off Shopify storefronts. Customers connect an agent to their account on Shop, Shopify's orders-and-tracking platform, and use one-click Shop Pay to check out inside the conversation, with orders funneled back to the Shop app. Shopify markets the feature as a "personal shopper" that can be given rules such as budget guardrails or fast-shipping-only filters, and buyers must approve each purchase. In September, Shopify announced a partnership letting Meta's Muse agent browse and buy on a person's behalf; now any agent can do so without a partnership integration.
The Verge AI · 🔥 0 - Critic: OpenAI and Anthropic ARR figures are invented numbers
An opinion piece argues that the widely repeated "annualized run rate" revenue figures for OpenAI and Anthropic are made-up numbers built on undefined periods and inconsistent gross-versus-net accounting. It cites reporting that OpenAI's September figure of $70 billion was later put at $50 billion after the total was inflated by grossing up sales through cloud partners, while Anthropic recognizes gross revenue on such partner sales and derives its run rate from the prior 28 days of consumption revenue plus monthly subscriptions. The author also contrasts these figures with reported first-half 2026 revenue of $12.4 billion for OpenAI and $15.5 billion for Anthropic, saying annualized numbers mainly serve marketing, search traffic and investor narratives.
Where’s Your Ed At · 🔥 0 - AI reputation management: how ChatGPT, Perplexity and Gemini describe you
This guide, contributed by reputation firm Reputation Resolutions, argues that ChatGPT, Perplexity and Gemini do not invent reputations but repeat what they find in training data and live web retrieval, so fixing the cited sources is more effective than arguing with a chatbot. It cites ChatGPT reaching 900 million weekly active users in February 2026 and Google's Gemini app passing 750 million monthly users by the end of 2025, and says Wikipedia accounts for 47.9% of ChatGPT's top-10 cited sources, Reddit 46.7% for Perplexity, and Reddit, YouTube and LinkedIn 21%, 18.8% and 13% respectively in Google AI Overviews. The recommended process is a baseline audit that records each platform's answer and cited links, sorting claims into accurate, outdated, false or mistaken-identity categories, fixing false allegations first and outranking content that cannot be removed. The user figures and citation shares are the article's own claims and were not independently verified.
The Next Web · 🔥 0 - OPPO partners with Manus to bring agentic AI to Find X10 via ColorOS 17
OPPO announced on October 13 a collaboration with Manus to integrate the company's agentic capabilities into AI Mind Pilot's new Agent Mode on ColorOS 17, debuting on the OPPO Find X10 series. The mode is designed to go beyond question-and-answer: it interprets a user's end goal, breaks it into steps, picks tools and works directly with files and photos on the device, such as assembling flight, hotel and attraction bookings scattered across apps into a personalized itinerary. According to OPPO, local tasks like finding, organizing and renaming files run on on-device models, and cloud processing requires explicit user permission; the new Agent Mode initially supports both Manus and ChatGPT, with OPPO also offering free usage credits.
Manus · 🔥 0 - a16z Top 100 AI apps: few pay for AI, but the top 1% spend about $900 a month
Andreessen Horowitz published the seventh edition of its Top 100 Gen AI Apps ranking, adding observed US consumer card spending data from YipitData panels for the first time. It says nearly half of US consumers use AI, but only about 4.5% had an active paid subscription to ChatGPT, Gemini or Claude in August, and spending is extremely concentrated: the top 1% of payers account for nearly a fifth of observed spending at roughly $900 a month, versus about $25 for the typical paying user. ChatGPT leads web traffic at roughly double Gemini's and six times Claude's, with Claude now a clear third and roughly level with Gemini in paid US subscribers.
The Decoder · 🔥 0 - a16z adds revenue ranking to Top 100 Gen AI Apps; half of top earners miss usage lists
a16z published the 7th edition of its Top 100 Gen AI Apps list, adding for the first time a Top 50 consumer AI ranking by monthly revenue; the usage-based Web and App lists added only 11 new products, the smallest change in three years. Revenue figures come from YipitData, which tracks card and receipt data from over 12 million US users, ranked by personal card payments to AI products in August 2026, with mainly NSFW-oriented products excluded. Of the 50 revenue leaders, 29 do not appear on the usage lists, and only ChatGPT, Claude, Perplexity, Suno, Photoroom, Canva and Notion appear on all three; the piece profiles lesser-known earners including AI 3D platform Meshy, workflow tool n8n and Nous Research's open-source personal agent Hermes.
36氪 人工智能 · 🔥 0 - Windows-MCP: open-source MCP server lets AI agents control Windows apps and UI
CursorTouch/Windows-MCP, trending on GitHub, is an MIT-licensed MCP server that connects LLM agents to Windows so they can navigate files, control applications, interact with the UI and run QA tests, without relying on traditional computer vision or task-specific fine-tuned models. It is published on PyPI (runnable via uvx windows-mcp) and listed in the MCP Registry, supports Windows 7 through 11, offers a DOM mode for Chrome, Edge and Firefox, and exposes stdio, SSE and Streamable HTTP transports plus auth keys, IP allowlists and tool whitelisting. The README claims 2M+ users through Claude Desktop extensions and 0.2–0.5s per-action latency; both are self-reported and unverified.
GitHub Trending · Python · 🔥 0 - Heathrow taps Adobe agentic AI to rebuild website and track LLM visibility
Heathrow Airport said on 8 October that it is extending its partnership with Adobe and will rebuild its main website on Adobe CX Enterprise, an AI customer-experience system that uses AI agents; the site draws more than 50 million visitors a year. The airport will also use Adobe Experience Manager for personalised content aimed at passengers, local communities and government stakeholders, plus Adobe Brand Visibility to track how its content surfaces on ChatGPT, Microsoft Copilot and Perplexity. Neither company gave a launch date or disclosed the deal's value.
The Next Web · 🔥 0 - a16z: top 1% of paying users drive 19.5% of consumer AI revenue
a16z released the seventh edition of its Top 100 Gen AI Consumer Apps ranking, adding YipitData US credit-card spending data for the first time. It says nearly half of Americans have used AI but only 25% do so daily, and as of August just 4.5% of consumers paid for ChatGPT, Gemini or Claude. Spending is highly concentrated: the top 1% of paying users generate 19.5% of revenue and spend about $903 a month, versus $25 for the median payer.
36氪 人工智能 · 🔥 0 - Getty Images delisted from NYSE as generative AI erodes stock-photo business
NYSE suspended trading in Getty Images' Class A shares on Sept 29 and began delisting procedures over an abnormally low price; Getty chose not to appeal and its stock moved to the OTC Pink market on Sept 30, keeping the ticker GETY. Reports put the last close at $0.042, down 47.3% in a single day, with market value shrinking to $17.42 million, while the company paid interest on its 2027 and 2028 notes on the final day of a grace period and is negotiating financing, including DIP loans, with lenders. Getty blames generative AI for collapsing the cost of imagery; its $3.7 billion merger with Shutterstock was terminated on June 30, and it is pivoting to display-licensing deals with Perplexity and OpenAI, though its record $981.3 million 2025 revenue still included one-off licensing payments.
科技新报 · 🔥 0 - Perplexity's open-weights Decider v1.1 hits OpenRouter at half the price
Perplexity has released pplx-decider-v1.1-27b, an updated open-weights multimodal decision model, now available on OpenRouter. It accepts text, JSON or image inputs and returns typed answers with probabilities. According to Perplexity Developers and OpenRouter, it costs $0.02 per million input tokens — half the price of v1 — with output free, and the company says it scores highest on the newly launched Hugging Face Decision Index 0.3 benchmark.
OpenRouter (X) · 🔥 0 - Fake ChatGPT, Gemini, Claude ads used in phishing to steal ad accounts
Security vendor Island published a report describing a phishing campaign in which attackers impersonate ChatGPT, Gemini, Claude, Perplexity, Manus and other major AI brands, advertising fake “AI marketing tools” that prompt agency staff, media buyers and manager-account admins to connect their Google account. Clicking “connect” launches a browser-in-the-browser (BitB) phishing page with a live human operator who fakes mistyped-password errors and chooses which MFA screen the victim sees, harvesting credentials and one-time codes. The lures target advertising accounts specifically because they carry stored payment methods, approved budgets and, for manager accounts, access to multiple client accounts; the campaign was still running and had not been disrupted when the researchers wrote the report.
TechRadar · 🔥 0 - a16z: Only 4.5% of US consumers pay for AI, while GPU rents keep rising
a16z's seventh consumer AI report finds that only 4.5% of US consumers have paid for a personal subscription to ChatGPT, Gemini or Claude, with spending following a steep power law: the top 1% of payers spend an average of $903 a month, more than the bottom 50% combined. Its second State of Markets report says compute demand is still running hot — A100 rental prices have not fallen and sit at or above early-2026 levels — while only about 30% of S&P 500 companies report quantifiable AI impact and roughly 2% have trackable metrics. The report also notes Meta Muse reportedly hit about 250,000 daily actives in its first week and 5 million downloads in under a month, and that ChatGPT Ads reached a $1 billion annualized revenue run rate in August 2026.
MIT科技评论中文 · 🔥 0
Experience and discussion from the community
Share my Perplexity experienceAsk about Perplexity
Nobody has shared their experience with Perplexity yet.