The OpenAI employee tender offer has closed at $7 billion, with the company buying back shares from its own workforce at a valuation of $852 billion, matching the figure set during its March 2026 funding round. Bloomberg first reported the completion of the deal.
What sets this transaction apart from prior employee liquidity events is how OpenAI funded it: according to CNBC, the company used its own cash rather than bringing in outside investors, keeping its capitalisation table free of new outside shareholders ahead of a potential public listing.
The OpenAI Employee Tender Offer in Context
This is not OpenAI’s first time running such a process. In October 2025, Thrive Capital, SoftBank, and others purchased $6.6 billion in employee shares at a valuation near $500 billion, according to beincrypto, citing Bloomberg. Before that, a 2023 tender tripled OpenAI’s valuation to $86 billion. Each successive offer has carried a sharply higher price tag and a sharply higher implied company value.
The pattern reflects a broader structural shift in Silicon Valley. With many high-growth technology companies staying private far longer than previous generations of startups, secondary tender processes have become a routine tool for retaining staff whose stock compensation would otherwise be illiquid for years.
The $852 billion valuation locks in the same figure from the March 2026 funding round that closed at $122 billion, up from the $110 billion figure announced earlier in February. That round was anchored by Amazon, NVIDIA, and SoftBank as strategic partners, with continued participation from Microsoft; SoftBank co-led alongside a16z, D. E. Shaw Ventures, MGX, TPG, and accounts advised by T. Rowe Price Associates. In a first, OpenAI also raised $3 billion from individual investors through bank channels during that round.
IPO Timing: The S-1 That May Sit on the Shelf
OpenAI submitted a confidential S-1 filing to the Securities and Exchange Commission (SEC) on 8 June 2026, working with Goldman Sachs and Morgan Stanley on the process, per CNBC. The company was characteristically candid about its own ambivalence. In its announcement of the filing, it stated: ‘We expect it to leak so we’re just announcing it. We have not decided on timing yet; it may be a while because there are things we want to do that are likely easier as a private company.’
A completed self-funded tender reinforces that equivocation. If an IPO were imminent, sourcing liquidity from the public markets would render the internal buyback redundant. Instead, the structure suggests OpenAI is managing employee pressure while it waits for conditions, internally and externally, to improve.
Those conditions are not uniformly favourable. The Wall Street Journal reported in April that OpenAI missed internal financial goals. Sam Altman, OpenAI’s chief executive, acknowledged the same trajectory last month, writing: ‘we did not have our best 12 months ever, which is mostly my fault, but we are about to have our best 12 months to date.’
The competitive pressure from Anthropic adds another layer of complexity. Anthropic filed its own confidential S-1 with the SEC on 1 June 2026, eight days before OpenAI’s submission, according to ActuIA. The rival lab was reportedly profitable earlier this year. Firms preparing for a public debut typically want their best numbers in the window investors will scrutinise most closely, and OpenAI may be calculating that it needs more time to get there.
On the operational side, the picture from OpenAI’s own announcements is more encouraging. The company’s enterprise business now accounts for more than 40% of revenue and is on track to reach parity with consumer revenue by the end of 2026. Its APIs process more than 15 billion tokens per minute. Codex, its coding assistant, serves over 2 million weekly users, a figure that is up fivefold in the past three months, with usage growing more than 70% month on month.
Whether that momentum translates into public-market readiness is the question the tender buys OpenAI time to answer. Anthropic’s IPO timeline, whenever it crystallises, may ultimately force the issue.
