Airtable is getting into the AI ​​agent game with Superagent

It may seem crazy to some, launching a brand new product line while your main business loses two-thirds of its paper value. But Hui Liu, founder and CEO of Pneumatic tablesuggests that this is the wisest thing he could do.

The company that investors valued at $11.7 billion during the zero interest rate period of 2021 is now trading on secondary markets for about $4 billion. But Airtable has raised a total of $1.4 billion, and Liu says the company still has half of that in the bank while it “sheds cash.” The valuation collapse affected investor returns and employee stock options, but did not undermine the business itself.

Leo’s response is to take off Super agentthe AI ​​agent he proposes could eventually outperform Airtable itself. It’s Airtable’s first standalone product in its 13-year history, and it captures the direction the company is headed and the reality of today’s AI moment: Every serious software player is racing to prove they can deliver to customers.

To understand what makes this move particularly interesting, consider what Airtable is: a no-code platform that democratizes building apps. It’s basically a super-fast database that allows anyone to create custom software tailored to their workflow. The company now employs more than 700 people and serves more than 500,000 organizations, including 80% of the Fortune 100. This is not a struggling startup but instead a mature company that is betting its future on a new architecture.

Superagent represents Liu’s bet on “multi-agent coordination” — a system where you ask a question and you don’t get a single AI assistant fumbling through sequential tasks, but rather a coordination agent deploying specialists working in parallel. “You’re not asking for AI,” Liu explains. “You’re organizing a team.”

Here’s how it works: When you ask Superagent about expanding your sports brand to Europe (example from Liu), the system first builds a research plan, identifying what needs to be investigated and showing you dimensions you didn’t think to ask about. Then you deploy specialized agents in parallel – one investigates financial matters, another analyzes competitive positions, and another reviews management and news. Finally, it brings everything together into a final product.

The output is not a wall of text. It’s an interactive market analysis that includes demographic details, a visually defined competitive presence, and expansion timelines that you can filter and explore. “What if everyone could get a New York Times-quality data visualization designed for every task they did,” Liu told me over Zoom last week. “This would have been unfathomable ten years ago, or five years ago, where you weren’t getting this quality of output — you were just getting the text. But to now be able to get really rich, high-quality interactive output as a virtual format, I think that’s a game-changer.”

The distinction Liu draws between Superagent and competitors is a technical one. He name-checked Anthropic’s AI agent Claude and Manus (a new entrant into the field of AI research). Acquired by Meta) as the only two products with a “generally capable, long-lasting, and truly nimble agent structure.” He argues that most other so-called agents are just “LLM-powered workflows” — pre-defined steps with a mix of AI calls, and not true autonomous agents that can course correct and undo.

It’s a fine line in a market where everyone is suddenly launching AI agents. OpenAI began in 2025 by launching new tools for building agents, while Notion, Harvey, and hundreds of other companies have since added agent functionality. In a market awash in proxy claims, Liu’s insistence that the super-agent is different will need to prove itself in practice.

In a blog post announcing the product, Liu gives examples of what Superagent can do. Ask him to evaluate Google as a three-year investment opportunity, he writes, and you’ll get a structured assessment with citations to earnings calls, analysis of the defense against OpenAI and Anthropic, and risk factors he didn’t take into account. Ask him to walk you through Wells Fargo’s AI strategy before you pitch it, and you’ll get a sense of their regulatory stance, their recent investments in AI, and the specific pain points your product addresses. The system pulls from premium data sources like FactSet, Crunchbase, SEC filings, and earnings transcripts.

The move culminates a transformation for Airtable, which Liu repositions as a “native AI platform.” Last fall, the company appointed David Azose, former engineering director of ChatGPT business products at OpenAI, to the position of CTO. At the same time, it It took over Deep Sky (formerly Gradient), an AI startup that has raised $40 million. Superagent will operate semi-independently from Airtable, which is run by DeepSky’s founding trio.

Pricing was still being negotiated as of last week, but it seemed poised to follow the emerging playbook for AI products: $20 per month per entry-level user, up to $200 for premium users, with generous inference credits. “We’re not trying to improve profit margin right now,” Liu says.

It remains to be seen whether Superagent will become the trillion-dollar market that Liu envisions or a big bet that doesn’t pan out. Competition is not trivial, and the distinction Liu draws between “real agents” and the rest of the business might not matter to clients if others were able to achieve sufficient results faster and at a lower cost.

But for a CEO whose company lost $7.7 billion on a paper valuation while retaining most of its actual capital, the move shows a willingness to bet on the future rather than protect the present. In fact, Liu reframed the previous valuation pressure as a hiring advantage, telling employees they were getting “stock that is actually more attractively priced than the $11 billion valuation” and a significant upside if his bets paid off. It has the capital for strategic acquisitions and does not need to raise another round.

When asked if he thinks Superagent is ultimately the bigger opportunity, Liu shrugs his shoulders — he’s not ruling it out. “Airtable is likely to be bigger, at least in the near term, than any new products we’re doing, including Superagent,” he says. “But I also like being able to bet on Superagent. Optionality is nice.”

This is Liu’s version of what he calls wartime leadership – a term he admits he once avoided as unnecessarily violent but now embraces as appropriate. “Being very adaptable is the most creative way to manage things right now,” he says. “It’s also the most exciting way to do things,” he quickly adds.