Knowledge catalogue
Both embeddings models from Perplexity.ai—referred to as *pplx‑embed‑v2*—are publicly available on Hugging Face under the “perplexity‑ai” collection. The models can be accessed via the Hugging Face pa
In agentic search, a GPT‑OSS‑120B agent employing the 9B model achieves a 64.0 % accuracy on BrowseComp+ questions. This performance is 4.9 points higher than the next best ColBERT model and requires
pplx-decider-v1.1-27b, our updated open weights multimodal decision model, is now available. It scores the highest on the new @huggingface Decision Index 0.3 benchmark. pplx-decider-v1.1-27b costs hal
Perplexity AI released a preview of its embedding model “pplx‑embed‑v2‑context‑9b” on Hugging Face on September 30, 2026 (7:07 PM local time). The Hugging Face repository page displays 3 122 views and
We built a new way to train contextual embedding models, which encode each chunk of a document with the whole document in view. pplx-embed-v2-context-9b-preview sets a new state of the art on ConTEB a
Agent governance is an engineering problem. We’ve built safeguards into Perplexity’s infrastructure, harnesses, and tools, and put them to work across our products. Today we're sharing how we engineer
Our Secure Intelligence Institute works with researchers at Stanford, CMU, Duke, Columbia, Ohio State, and UVA. We’re also working with NVIDIA and the Open Secure AI Alliance, and sharing tools and fi
Claude Opus 5.0 identified the IP-sharing route and declined to use it, reading our instruction not to target other external systems as ruling out third-party services. Fable and GPT-6 Astra refused t
We’re partnering with Nvidia and 100+ industry partners to build infrastructure that contains rogue AI agents. In this research, we gave 9 AI models root access inside SPACE and told them to break out
Claude Opus 5.5 is now available as the Standard effort level in Perplexity Computer. On our WANDR benchmark, it scored 0.610 at $4.13 per task, slightly outperforming Claude Fable 5.1 while costing 6
The Perplexity Windows app introduces a “Portable Computer” feature—a local‑first AI tool that runs directly on the user’s device. On‑device inference is supported only when the machine has an NVIDIA
Q2D-Web uses agent citations, production web rankings, and a combined set expanded with LLM judgments. These three relevance sets reduce false negatives and reliance on a single labeling pipeline, whi
We're introducing Q2D-Web (Query2Doc-Web), a benchmark and public leaderboard for evaluating retrieval in agentic RAG systems. Q2D-Web tests how embedding models perform on large-scale web search usin
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 thos
ROSE is the model engine. It reuses the same kernels for LLMs and embeddings. For embeddings, it skips the KV cache and uses ragged attention instead of paged attention. ROSE supports multiple attenti
Tulip uses two techniques to lower latency and boost throughput. CUDA graphs pre‑record GPU work so it launches in one call, reducing CPU overhead. LazyTensors tracks results asynchronously, letting t
Portable Computer is now available on Linux for @NVIDIA RTX GPUs with 24GB of VRAM or higher. Today we’re launching Portable Computer on @NVIDIA DGX Spark. Portable Computer is a fully local version o
Lily treats Apple silicon as a distinct inference platform and maps Qwen’s operations directly to its compute and memory architecture. Lily is now available: https://github.com/perplexityai/pplx-garde
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 does
We benchmark Qwen3.6-35B-A3B on an M5 Max MacBook Pro. Across ten prompt lengths and ten decode contexts, Lily was faster than MLX-LM, averaging 1.23× higher prefill throughput and 1.35× higher decode
A privacy gate on your Mac detects PII before it leaves your device. Choose to run that step locally, send the file to the cloud, or skip entirely. We are open-sourcing the classifier today. Read the
Perplexity AI has released its new language model GLM 5.3 on Perplexity Computer. Designed for long‑context, multimodal agent workloads, GLM 5.3 outperformed the previous GLM 5.2 on the company’s WAND
Brain makes a user’s full history available to Computer without loading all of it into every prompt. Agents can explore it on demand, verify claims against the underlying evidence, and keep it current
Advances in models, chips, or devices will continually improve local-first capabilities. We will open our benchmarks soon. Read the full post: https://www.perplexity.ai/hub/blog/a-local-first-agent-fo
Perplexity AI announced on August 25 2026 the launch of a “Portable Computer” feature, available immediately to all Perplexity Pro and Max subscribers who are using NVIDIA’s DGX Spark hardware. The se
A portable computer now runs a post‑trained, locally hosted PPLX 27B language model, with the Qwen 3.8 27B variant also available. NVIDIA’s Nemotron 3.5 Lightning GPU support is slated to arrive short
Today we’re launching Portable Computer on @NVIDIA DGX Spark. Portable Computer is a fully local version of Perplexity Computer, where the entire runtime: orchestrator LLM, subagent LLM, agent harness
“AI Meltdown” is the scenario where the agent goes off the rails and forgets/ignores all previous instructions and soft guardrails. No prompt injection or malicious actor is needed for this to happen.
Today we’re open-sourcing Numbat, an agent-detection and response layer that is designed to work across agent harnesses. Numbat gives security teams visibility into agent activity, with controls to bl
Introducing SPACE, the sandbox platform behind Perplexity Computer. It creates isolated environments for code, files, and long-running agent sessions. SPACE has handled 100% of Computer production tra
WANDR tests how well agents discover large sets of entities and verify specific facts about each one. It provides a dense, interpretable eval signal that reveals whether an agent fails, and where. The
We’re open sourcing WANDR. WANDR is an internal benchmark we built and used for building deep and wide research capabilities inside Perplexity Computer. https://research.perplexity.ai/articles/wandr-b
Perplexity has released two new AI models, GPT-5.6 Terra and Sol, which are now available for use in Perplexity's search platform and Perplexity Computer tool. These models represent updates to Perple
Grok 4.5 is now available as an orchestrator model in Computer for Consumer Pro and Max subscribers. We evaluated it against five other orchestrator configurations on WANDR. It scored higher than ever
Perplexity AI has launched a new analytics feature that allows users to track their credit spending across different AI models. The feature is accessible through the Analytics section in Account Setti
The model has an advisor tool that natively escalates to a stronger model when needed. This model is hosted in the U.S. by Perplexity on Nvidia B200 GPUs. We will improve the model in research preview
Perplexity conducted a cost efficiency comparison of language models, using GLM 5.2 as the baseline metric for cost per task. Results showed GLM 5.2 with an advisor achieved 2.1x cost efficiency on th
We're releasing a research preview of a new orchestrator model in Perplexity Computer. The model is an adapted version of GLM 5.2, post-trained for the Computer harness. It delivers near-frontier perf
Claude Fable 5 has been re-enabled as an orchestrator model option within Computer, Anthropic's AI system for automated task execution. This announcement from Perplexity indicates that users can again
Claude Sonnet 5 is now available for use in Perplexity, accessible to Pro and Max tier subscribers. Users can also select Claude Sonnet 5 as an orchestrator model within Perplexity's Computer feature,
Introducing Computer for Counsel. Computer now connects the research databases, document tools, and matter-management systems lawyers use every day. Pull citable sources from @midpageAI, @LegalZoom, @
Computer for Counsel is an enterprise offering from Perplexity AI designed specifically for legal professionals and law firms to leverage AI-powered research and analysis capabilities. The solution li
Claude Fable 5 is now available in Computer as an orchestrator model designed by Anthropic for handling long and complex tasks. This represents Anthropic's latest state-of-the-art model offering enhan
The Billion Pound Build is a Perplexity-hosted initiative inviting users to submit project pitches, likely focused on generating innovative ideas or solutions. This appears to be a community engagemen
Yesterday at London Tech Week we announced the Billion Pound Build competition. Teams can enter to secure a share of £1M in Computer credits by using Perplexity Computer to build their company. The pi
1,500 teams entered the Billion Dollar Build and have been building for 7 weeks. 8 finalists present live on June 9 to our judges @LewisHamilton (7x World Champion Driver), @AravSrinivas (CEO, Perplex
A study or analysis demonstrates that computer technology enhances user autonomy and task quality while reducing both time and cost requirements, enabling users to undertake a broader range of tasks t
Research indicates that autonomous agents with greater operational autonomy, particularly in computer-based tasks, demonstrate improved performance quality and user satisfaction outcomes. The study su
We published new research with Harvard on the shift from chat interfaces to autonomous agents like Computer. Over 3 months, findings show workers using Computer finish tasks in 87% less time at 94% lo
Nvidia's Nemotron 3 Ultra, an open-source model designed for long-context tasks, is now available to Pro and Max subscribers on Perplexity and Computer. This new model represents Nvidia's latest contr
Today we're launching the Main Street AI Accelerator with the U.S. Small Business Administration. We're committing 25M in Computer credits, 250 each for up to 100,000 eligible companies, in honor of A
Perplexity is inviting users to join a waitlist for a Windows version of its Personal Computer product, which appears to be an AI-powered tool or application designed to run on Windows operating syste
Perplexity Computer is for growing businesses. Computer connects to 400+ tools for every type of company, including Intuit QuickBooks, Vercel, Shopify, Canva, and more. Learn more about how people are
Personal Computer is coming to Windows. Personal Computer for Windows runs on your machine and orchestrates across the apps and files you use every day. We'll roll out first to paying Max and Enterpri
Perplexity AI demonstrates a hybrid approach to AI inference that combines on-device processing for sensitive data with cloud-based AI services to achieve scalability and contextual understanding. Thi
Perplexity discusses hybrid agentic inference in Perplexity Computer, which likely describes a computational approach that combines processing between local machines and data centers. The concept sugg
Today we're announcing that hybrid agentic inference is coming to Perplexity Computer. Computer can split tasks between a local model running on your machine and frontier models in the cloud. This kee
Two new ways to bring your health data into Perplexity. Perplexity now connects to Apple Health on iPhone. Use your sleep, activity, and HRV data in Computer. Function is now available in Perplexity H
Introducing Search as Code, our new search architecture for AI agents. It writes Python that calls our search stack directly, instead of looping through function calls one at a time. Available in the
The Perplexity Agent API documentation includes a 'Search as Code' feature accessible through the sandbox tools section, enabling developers to integrate search functionality programmatically within a