a daily news desk
Deployments

OpenAI files confidential S-1, putting both frontier-model vendors on a public-market track

A week after Anthropic's draft S-1, OpenAI's own filing creates a $3.6T AI IPO pipeline alongside SpaceX — and a first look at the financials underneath the agent stack.

By Ben Krauss · Daily desk lead ·

OpenAI submitted a confidential draft S-1 to the SEC on June 8, exactly one week after Anthropic filed its own. Both leading frontier-model vendors are now on parallel public-market tracks, with Goldman Sachs and Morgan Stanley leading OpenAI’s process and Bloomberg reporting a potential fall 2026 listing window.

OpenAI didn’t wait for the leak. The company published a Rule 135 disclosure on its own site: “We recently submitted a confidential S-1. We expect it to leak so we’re just announcing it.” A second line in the same post hedges the timeline: “it may be a while because there are things we want to do that are likely easier as a private company.” That’s narrative management in two sentences. Pre-empt the scoop, then condition expectations.

The numbers underneath are where the dual-filing posture gets interesting. OpenAI’s last mark was $852 billion, set in March’s $122 billion round. Anthropic has since printed a paper valuation of $965 billion on a $47 billion revenue run rate that TechCrunch attributes to enterprise adoption of Claude for coding and agentic workflows. ChatGPT, per CNBC, now serves more than 900 million weekly active users. Two very different shapes of business, about to be priced against each other on the same tape.

Add SpaceX, freshly merged with xAI and already on its roadshow at roughly $1.75–$1.8 trillion, and Bloomberg’s combined IPO pipeline lands near $3.6 trillion. Bankers have reportedly told both OpenAI and Anthropic the same thing: there’s a first-mover advantage in defining how public investors categorize the sector. SpaceX’s own filing already names OpenAI, Anthropic and Google as key AI competitors, which tells you how the roadshow framing is being staged.

The structural consequence is that the S-1 itself becomes the document. Inference margins, compute commitments, customer concentration: everything the API buyer currently can’t see is about to become a filing exhibit. The category gets defined by whoever files the cleanest one.

Sources