Parallel, the AI-first web platform from ex-Twitter CEO, aims to outperform GPT-5

The former Twitter CEO who was ousted when Elon Musk took over the platform in 2022 is making headlines again but this time, for launching a bold new startup aimed at reshaping the internet for the age of artificial intelligence.

Agrawal’s new venture, Parallel Web Systems Inc., or simply Parallel, is building a platform designed to help AI systems search, understand, and use the web more efficiently.

What is Parallel?

Founded in 2023 and based in Palo Alto, Parallel is a cloud-based platform that allows AI systems to do large-scale web research the kind of deep, structured searching that goes far beyond simple queries.

Instead of clicking links or scrolling pages like humans do, AI tools using Parallel can ask complex questions and instantly comb through vast amounts of data from documents and databases to real-time web content.

Agrawal says this is necessary because the internet, as it stands today, was built for people but the next era belongs to artificial intelligence.

“The web is humanity’s memory,” Agrawal wrote in a blog post. “AI is becoming the web's primary user.”

What makes it different?

Parallel is creating what it calls a “Programmatic Web”, a new kind of internet that works for AIs, not around them.

Here’s what that means:

AI-first infrastructure: Instead of static pages, Parallel provides AI systems with data, compute power, and reasoning tools to deliver useful answers and insights.

Smart interfaces: AIs tell the system what they need, and Parallel figures out how to deliver it.

Source transparency: Every bit of information comes with proper credit, so content creators get recognition.

Incentives for openness: Parallel plans to reward contributors through open economic models, making the system fairer for everyone.

A new API that beats GPT-5?

One of the biggest reveals from Agrawal’s announcement is the launch of Parallel’s Deep Research API, which the company claims has already outperformed humans and leading AI models like GPT-5 on key research tasks.

Agrawal shared that fast-growing AI startups and even a public company are already using Parallel to power research, automation, and coding tools.

Backed by big names

Parallel has quickly gained the attention of top-tier investors, securing $30 million in funding from firms like Khosla Ventures, Index Ventures, and First Round Capital. The company now has a 25-person team working out of its Palo Alto office.