Parallel Web Systems

Introducing the Parallel CLI: the most intuitive way for agents to access high-quality data from the open web. - Search: Search the web in natural language - Extract: Get clean page contents from any URL - Deep Research: Run deep research on open-ended questions - Enrichment: Enrich CSV or JSON data with AI-powered web research Read the blog and get started with a single command in your terminal: parallel.ai/blog/parallel-…

@jaga1520 💥

@p0 Game changer!

You can now create stateful web research agents with the Parallel Task API. Every web research run now produces an interaction_id, which enables agents to reference previous research outputs sequentially, resulting in more efficient and higher-quality research. Try interactions in the new developer platform showcase: platform.parallel.ai/play/interacti…

@TinaJiang20 so much easier to ask follow ups now

@p0 cool to be able to now have threaded conversations!

@travers00 it really does!

@p0 this unlocks a ton

Previously, Task runs were isolated sessions. Now they can be chained together to perform complex, multi-turn knowledge work on the web. Importantly, each new turn can have different configurations: different Task Processors and input/output schema, allowing for more flexible, composable, and sophisticated deep research use cases. For example, starting with an open-ended question to fan out on a topic, before refining the scope of the next query to focus in on a specific part discovered in the first turn. To learn more and get started, read today’s developer blog: parallel.ai/blog/task-api-…

MPP comes with a directory of 100+ services for your agent to use & pay for autonomously on Tempo: • @alchemy — blockchain data across 100+ chains • @alliumlabs — onchain analytics • @browserbase — headless browsers • @dune — onchain SQL queries • @fal — image, video, and audio generation • @merit_systems — phone calls, travel, email, & more • @p0 — web search and deep research • @postalform — mail delivery via USPS • @prospectbutcher — sandwich ordering • ... and many more No API keys. No accounts. mpp.dev/services

The biggest infrastructure decisions aren't made when the market is obvious. They're made when almost no one else sees the shift coming. Two years ago, @paraga started building search infrastructure — not for humans, but for AI agents ( @p0). The customers he was building for hadn't fully arrived yet. He built anyway. In this episode of Founded & Funded, Madrona Venture Partner @JTurow sat down with Parag and @nikitabase (who led @neondatabase through an acquisition by @Databricks ) at the IA Summit. Here's what struck most during their conversation: 1) When your end customer changes, everything in your stack has to change with it 2) Neon saw agents create databases 4X faster than all human developers — in one channel 3) Both founders had to decide: lean in or get left behind 4) An AI-native search index requires a different crawl strategy, a different ranking model, and a different relationship to what's already in model memory Watch the full episode: YouTube: ordnl.link/wOr6321 Spotify: ordnl.link/t4HQjim Apple: ordnl.link/bSlj5Gk Amazon: ordnl.link/AYQFsu9 Read the transcript: ordnl.link/FFZR3xW

Two founders. Two companies. One shared conviction. Founders who wait for certainty don't miss the bet. They miss the whole wave. @Paraga built Parallel Web Systems ( @p0) from scratch around the idea that an AI was going to sit between infrastructure and the end customer — and that every product decision made for a human on a mobile device would have to be remade. @nikitabase watched it happen in real time at @neondatabase, and put it plainly: when a disruption hits and you don't move, you end up selling pagers while the world moves to iPhones. In the latest episode of Founded & Funded, Madrona Venture Partner @JTurow sat down with both of these founders live at the IA Summit to talk about what it actually looks like to bet the company on a customer that's still arriving. What comes out of the conversation: +Agents were spinning up Neon databases at 4X the rate of human developers — from a single platform — before Nikita had a plan for what to do about it +Parag built a new search index from scratch because existing ones were designed for humans, and that assumption breaks entirely when the customer is a model +Neon shipped a production MCP server one week after Anthropic launched the protocol, before anyone knew if it mattered, and it mattered two months later +Parag's hiring framework for teams taking on real risk: two types of people, risk adders and people who can channel that risk into a concentrated bet — you need both or you get diffusion, not a bet +Nikita's take on agent autonomy: build two-way doors, fork environments, run agents in parallel, pick the winner Watch the full episode: YouTube: ordnl.link/wOr6321 Spotify: ordnl.link/t4HQjim Apple: ordnl.link/bSlj5Gk Amazon: ordnl.link/AYQFsu9 Read the transcript: ordnl.link/FFZR3xW

The largest buyers of software will soon be agents autonomously purchasing tools at runtime, paying directly for usage with no account setup. Machine Payments Protocol is the payment layer that makes that possible, and Parallel is live on day one. So fun to work with our friends at @stripe and @tempo on this!

Imagine that instead of agents (or their human handlers) spinning up accounts and API keys for every service they use, they just post an API request with its payment inline. That's the reality we now live in with @tempo and @mpp

Parallel is built for a world where agents are the primary users of the web. Agents can now autonomously pay per API call for web access via Parallel using MPP. parallelmpp.dev

Demo time! 🤖💸🎉 Here's @stevekaliski, @BackseatVC, and their agent (with a wallet) throwing together an actual birthday party. To accomplish this fun task, the agent autonomously researches, requests, and pays multiple businesses on Stripe via MPP. Note: No hands on keyboard

Tempo Mainnet is live! Starting today, anyone can build on Tempo through our public RPC endpoints. Alongside mainnet, we’re introducing the Machine Payments Protocol, an open standard for machine payments.

MPP made it dead simple for @p0 to accept global stablecoin payments from agents and have funds settle as USD in our existing Stripe billing stack

Summarizing the launch article from @tempo using one of the MPP integration, Parallel (@p0) > Tempo is a high-performance blockchain network purpose-built for stablecoins and payments. > It delivers sub-second finality and near-zero transaction costs, making it the foundation for the next generation of financial infrastructure. @dwr 🫡

@tempo @mpp @stripe 🌐🤝

Agent payments will soon overtake human payments on the internet. The Machine Payments Protocol (@mpp) is a new open standard co-authored by @stripe and @tempo. It’s designed to be extensible and payment-method agnostic, already supporting stablecoins, cards, and more.

Parallel exists for AI to access the web natively, to power high-accuracy, superhuman-scale knowledge work, but also to ensure information on the web remains open in the AI era. Native agent payments via MPP are a step towards that vision. It helps lay the foundation of an open market for content and data, where information can be exchanged in ways that better sustain its creators. Where web data transacts programmatically between buyers and sellers. To learn more about MPP and Parallel, read today’s blog: parallel.ai/blog/tempo-str…

For developers, getting started on MPP is as simple as giving the following prompt to your preferred agent: Read tempo.xyz/SKILL.md to set up tempo Then read #agents" target="_blank" rel="nofollow noopener">parallelmpp.dev/#agents to set up parallel Get started docs.parallel.ai/integrations/t…

Parallel is live on @tempo. Agents can now discover and pay for web search, content extraction, and multi-hop research per use— no account setup and no API key— via the Machine Payments Protocol (MPP).