OpenAI Financials: October 2026 Update

OpenAI enters October with its revenue run-rate jumping sharply to nearly $70B annualized and a reported fundraise in motion that would reset its valuation well above the $852B post-money mark set in March.

Key points

• Axios, citing internal figures, reported OpenAI’s annualized revenue is approaching $70B as of late September, up from $40B at the end of July.
• OpenAI is seeking at least $30B in new capital at roughly a $1.4T valuation, according to Bloomberg, making this potentially the company’s final private round before a public listing.
• Executives confirmed an IPO will wait, citing the need to prioritize AI safety and noting the company is well-capitalized.
• At DevDay (September 29), OpenAI launched Dots always-on agents, a $500/month subscription tier, ChatGPT Spaces, a plug-in Marketplace, and a new ChatGPT-centered office suite that puts it in direct competition with Microsoft.
• GPT-6.1 Astra was pulled from a planned October release after failing internal safety evaluations; GPT-6.1 Sol shipped separately, nearly matching Astra at lower cost.

What does the revenue jump tell us?

The move from $40B to nearly $70B annualized in roughly two months reflects a roughly 75% step-up, driven by a combination of consumer and enterprise adoption. OpenAI cut API prices by more than 50% in September with the GPT-6 Sol and Luna launches, a move that typically expands volume while compressing per-unit margin. The $500/month Dots tier runs the other direction, targeting power users and enterprise budgets where margins are higher. If both vectors hold, the mix shift toward higher-margin agentic subscriptions could support the revenue trajectory even as cheaper inference tiers grow.

For context, Sacra pegged Anthropic’s run-rate at $65B as of July 31, which would put OpenAI back ahead of its closest private-market rival on revenue, at least by the latest reported figures.

Is a $1.4T valuation grounded in the numbers?

The reported ask is roughly 20x the approaching $70B run-rate. That is a compression from the March round’s implied multiple (852B / 40B = ~21x) but still aggressive in absolute terms. For reference, the public hyperscalers trade at far lower revenue multiples: Microsoft (MSFT) closed October 1 at $515, Amazon (AMZN) at $248.92, and Alphabet (GOOGL) at $339.24. The AI-specific premium in OpenAI’s multiple reflects growth rate, platform ambition, and scarcity of public comparables, not current profitability.

The March 2026 round at $852B remains the last closed and verified valuation. The $1.4T figure is a reported ask for a round not yet closed, so the snapshot carries $852B until a deal is confirmed.

IPO watch: still waiting

OpenAI management reiterated in late September that a public listing is not imminent. The framing is that a new $30B private raise serves as a bridge, described in Bloomberg’s reporting as likely the final private round. That framing has appeared before and may shift again, but for now the IPO timeline remains open-ended. The structural conversion to OpenAI Group PBC, completed in late 2025, removed the main governance obstacle to a public offering. What remains is management’s stated preference to stay private while the safety and product roadmap matures.

Product velocity is real: DevDay produced Dots, Spaces, the Marketplace, Codex updates, the office suite, and GPT-6.1 Sol in a single week. The delay of GPT-6.1 Astra for safety reasons, while a short-term miss, signals that the company is applying meaningful friction to frontier releases, which is the kind of governance story that institutional public-market investors typically want to see before an IPO.

How do the comparables look?

Among private AI labs, Anthropic holds the higher Sacra valuation at $965B (as of July 10) on a $65B run-rate. OpenAI’s $852B post-money from March is now paired with a run-rate that has surpassed Anthropic’s last reported figure, which values OpenAI at a lower multiple of revenue than its closest peer, assuming the $852B figure is used. A closed $1.4T round would flip that relationship sharply. Among public names, the hyperscalers continue to trade at multiples that make both private labs look richly valued, though the growth differential is substantial.