OpenAI confidentially filed for an IPO targeting a September 2026 listing at a valuation expected to top $1 trillion, with Goldman Sachs and Morgan Stanley as co-advisors. The S-1 will be the first public look at OpenAI's audited financials. For developers and founders building on the platform, the questions that matter aren't about the stock -- they're about API pricing stability, platform lock-in risk, and what public-company incentives do to an infrastructure provider.
I've been watching this since OpenAI raised $122 billion in March and confirmed $2 billion in monthly revenue. The confidential filing moving today tells me the governance restructuring is complete and September is a real target window. Here's what it means for anyone who has OpenAI in their stack.
What Did OpenAI Actually File?
OpenAI submitted a confidential S-1 to the SEC on May 22, 2026 -- the first formal step toward a September public listing at a $1T+ valuation. Goldman Sachs and Morgan Stanley are co-advisors. A confidential filing allows private SEC review before public disclosure. The public S-1 drops weeks before the roadshow, likely July or August.
The $1 trillion target comes from the March 2026 funding round that established an $852 billion post-money valuation after raising $122 billion -- the largest private capital raise in history. The listing window Goldman and Morgan Stanley are targeting runs from Labor Day to Thanksgiving 2026, giving roughly four to six months of runway after the confidential filing processes through SEC review.
One major obstacle cleared just before the filing: OpenAI won a jury verdict against Elon Musk's structural claims, with the court ruling his arguments were time-barred. That removes the most significant legal overhang that had threatened OpenAI's governance ahead of a public offering. The governance path -- OpenAI's conversion from a capped-profit structure to a public benefit corporation -- was finalized earlier this year as a prerequisite for the IPO.
What Do the Financials Actually Say?
OpenAI hit $25 billion in annualized revenue by February 2026 but lost roughly $9 billion in 2025 -- spending $22B to earn $13.1B in revenue. The company is growing fast but is not profitable. CFO Sarah Friar publicly said OpenAI is "not ready to go public" due to spending commitments risk.
The $2B monthly revenue run-rate confirmed in the March 2026 funding round is the headline number that supports the trillion-dollar story. Enterprise is now the growth engine: businesses represent more than 40% of revenue, OpenAI has over 1 million business customers and 9 million paying business users, and enterprise is on track to reach revenue parity with consumer by end of 2026. Enterprise annual contract values average around $561,000 according to SpendHound benchmark data -- this is contract revenue, not API consumption, and it's what makes the IPO model work.
The structural risk that will dominate the S-1 is Microsoft. OpenAI's pre-IPO investor documents explicitly identify Microsoft as supplying "a substantial portion of our financing and compute." Microsoft is also a significant shareholder with a complex IP-sharing and revenue-participation agreement. If that relationship changes post-IPO -- either through Microsoft reducing compute commitments or renegotiating terms as a public company competes more directly with Azure AI -- the downstream effects on API capacity and pricing are real. The risk factor disclosure from CNBC's reporting makes this explicit before the full S-1 is even public.
Will Going Public Change API Pricing?
API pricing is unlikely to rise in the near term. OpenAI has been cutting prices consistently -- GPT-5.5 launched at $5.00 per million input tokens with a 50% Batch API discount available. Competitive pressure from Anthropic and Google keeps the floor in check. The bigger risk is multi-year Guaranteed Capacity lock-ins, not headline rate increases.
Here's the actual current pricing picture. GPT-5.5 runs $5.00 per million input tokens and $30.00 per million output tokens at standard rates. The previous flagship GPT-5.4 is $2.50/$15.00. For high-volume workloads, the Batch API processes requests asynchronously within 24 hours at 50% off: GPT-5.5 drops to $2.50/$10.00 per million tokens. Prompt caching cuts cached input further to $1.25/million on GPT-5.5. These are aggressive rates -- not the behavior of a company planning to raise prices post-IPO.
The mechanism to watch instead is the Guaranteed Capacity program. OpenAI is offering enterprise customers multi-year compute access deals ahead of the listing. Contracted revenue looks excellent in an S-1 -- it provides visibility that public-market analysts want, and it reduces churn risk. But multi-year commitments also create switching costs. The program is described as specifically designed to lock in the next generation of startups before they evaluate alternatives. Sign a two-year capacity deal at today's pricing and you get pricing certainty -- and reduced optionality if Anthropic or Google offers meaningfully better terms 18 months from now.
Want this built for your business?
Venti Scale builds AI automation systems for businesses that want results without the learning curve. One operator, AI-powered, full marketing stack.
See What We Build
How Does This Shift the Competitive Landscape?
Anthropic holds 40% of the enterprise LLM API market versus OpenAI's 27%, and surpassed OpenAI in enterprise adoption for the first time in April 2026. ChatGPT's overall market share dropped from 77.43% to 56.72% over the past year while Gemini tripled to 25.46%. The IPO is partly a credibility signal to enterprise procurement teams who need governance stability before signing multi-year contracts.
The market share data deserves unpacking. Anthropic's 40% figure comes from enterprise LLM API market share -- the segment of businesses choosing which model provider to build production applications on. That's the segment OpenAI needs to defend most as a public company, because enterprise contracts are the revenue base that justifies a $1T valuation. Consumer market share (ChatGPT subscriptions) is real but carries different margin characteristics. Enterprise API revenue is where the S-1 story has to hold up under analyst scrutiny.
Anthropic is simultaneously targeting an October 2026 IPO at a valuation approaching $900 billion -- potentially overlapping OpenAI's listing window. Both companies going public within months of each other creates a comparative data environment that analysts and builders will mine for information both companies have been unwilling to share: actual cost of compute per model call, actual API margin structures, and which growth rate is sustainable. Read both S-1s when they drop. That's the most useful competitive intelligence either company has ever had to publish.
Google's Gemini at 25.46% market share is the underrated story. Google can subsidize API pricing at a scale neither OpenAI nor Anthropic can match, because Gemini revenue flows into a company with $350B+ in annual revenue. Google's competitive response to two AI-native IPOs in the same quarter could include another round of pricing cuts or compute access incentives. Keep that in your model when planning infrastructure costs for 2027.
What's the Real Risk for Builders on OpenAI?
The biggest risks for builders aren't about stock price -- they're structural. Microsoft supplies a substantial portion of OpenAI's financing and compute per pre-IPO investor documents, and that dependency is hard to unwind. Model deprecations accelerate under public-company cost pressure. Multi-year Guaranteed Capacity programs shift the best pricing toward committed enterprise customers over pay-as-you-go builders.
Model deprecation is the risk most builders underestimate. Public companies retire costly legacy infrastructure faster than private ones, because every model generation that's still running costs compute and engineering time to maintain. OpenAI has deprecated multiple models faster than builders expected over the past two years -- the shift from GPT-4 to GPT-4o, then to GPT-5, required code updates and testing across thousands of applications. That pace typically accelerates under quarterly earnings pressure, not slows. If your application depends on specific model behavior in a current-generation model, build with deprecation timelines in mind from day one.
The Microsoft dependency is structurally complex in ways the S-1 will have to address. Per CNBC's reporting on OpenAI's pre-IPO risk disclosures, Microsoft controls a substantial share of compute capacity and has revenue participation in OpenAI API sales. If Microsoft's Azure AI products compete more directly with OpenAI's enterprise offerings post-IPO -- which is already happening with Microsoft Copilot -- the compute relationship gets complicated, not simpler. As a public company, OpenAI will be required to disclose when that relationship materially changes.
My honest take: the IPO is net neutral for most builders over the next 12 months. Pricing stays competitive because Anthropic and Google won't allow otherwise. The S-1 when public will contain the most useful information OpenAI has ever had to disclose -- read it. The Guaranteed Capacity lock-ins are worth being deliberate about before signing. Where it gets more interesting is 2027-2028, when a profitable-or-not OpenAI is making platform decisions under shareholder pressure. Having optionality in your AI infrastructure -- not being exclusively locked into one provider -- will matter more then than it does today.
FAQ
When is OpenAI's IPO expected to happen?
OpenAI is targeting a public listing between Labor Day and Thanksgiving 2026, following a confidential S-1 filing submitted to the SEC in May 2026. Goldman Sachs and Morgan Stanley are co-underwriters. The public S-1 becomes available weeks before the roadshow, likely in July or August 2026, and that's when audited financials will be public for the first time.
What is OpenAI's current valuation heading into the IPO?
OpenAI's March 2026 funding round established an $852 billion post-money valuation after raising $122 billion -- the largest private capital raise ever. The IPO is targeting a valuation above $1 trillion. OpenAI hit $25 billion in annualized revenue by February 2026, with $2 billion per month confirmed in the March funding disclosures, offset by a roughly $9 billion net loss in 2025.
Will OpenAI going public affect API pricing for developers?
API pricing is unlikely to increase in the near term. OpenAI has been consistently cutting prices -- GPT-5.5 runs $5.00 per million input tokens at standard rates, with 50% Batch API discounts available. Competitive pressure from Anthropic at 40% enterprise market share and Google keeps the pricing floor in check. The more relevant risk is multi-year Guaranteed Capacity lock-ins being marketed to enterprise customers ahead of the IPO.
Is it safe to keep building on OpenAI through the IPO?
For most use cases, yes. The IPO does not change API availability or reliability in the next 12 months. The structural risks worth monitoring: Microsoft compute dependence identified in pre-IPO risk documents, model deprecation pace accelerating under public-company cost pressure, and multi-year capacity programs shifting best pricing toward committed enterprise customers over pay-as-you-go builders. Build with provider optionality in mind for the longer run.
Want this built for your business?
Venti Scale builds AI automation systems for businesses that want results without the learning curve. One operator, AI-powered, full marketing stack.
See What We Build
