OpenAI Hits $2B Monthly Revenue, Eyes IPO at $852B Valuation After Record $122B Round

At the end of Q1 2026, OpenAI crossed a threshold that few companies in history have reached in under a decade of operation: $2 billion in monthly recurring revenue. For context, Salesforce — the enterprise software giant that pioneered the SaaS model — took more than 20 years to reach that run rate. OpenAI has built to that scale while simultaneously redefining its product, its governance structure, and its relationship with Microsoft.

The revenue milestone arrived alongside a $122 billion funding round — the largest single venture investment in history — valuing the company at $852 billion. The round was backed by Amazon, Nvidia, and a syndicate of sovereign wealth and institutional investors. If OpenAI proceeds with the IPO it is actively preparing, the offering would rank among the largest in history.

The Numbers Behind the Milestone

The $2B monthly revenue figure represents approximately $24 billion in annualized revenue — roughly on par with Adobe’s current annual revenue, and ahead of companies like Snowflake, Databricks, and most publicly-traded enterprise software firms. OpenAI crossed $1 billion in annual revenue in 2023; the acceleration from $1B/year to $24B/year in approximately three years is without precedent in enterprise software history.

The user base underlying that revenue is equally remarkable. OpenAI is approaching 1 billion weekly active users across ChatGPT and its API ecosystem — a scale that puts it in the company of Facebook (2012 vintage), WhatsApp (2020), and YouTube (current). Consumer AI went from a niche research curiosity to a near-ubiquitous utility in roughly 36 months.

The Q1 $122B round, which also included participation from Nvidia and others, was priced at $852B — a valuation that would make OpenAI the fourth most valuable US company by market cap if it went public at that price today, trailing only Apple, Nvidia, and Microsoft.

GPT-4o Retirement and the Model Stack Consolidation

On April 3, 2026 — just 48 hours before this report — OpenAI retired GPT-4o from all subscription plans and the free tier. The model that was, until very recently, the company’s primary consumer-facing offering has been replaced by GPT-5.x series models across paid plans and GPT-4.1 on the free tier.

The retirement marks a clean break from the architecture that defined OpenAI’s consumer product through 2024 and most of 2025. GPT-4.1, while positioned as a free-tier model, outperforms GPT-4o on most benchmarks — a sign of how rapidly the quality floor of the AI model stack has risen.

The consolidation around GPT-5.x for paid users and GPT-4.1 for free users simplifies a model menu that had grown complex and confusing for non-technical users. It also enables OpenAI to retire the significant computational infrastructure required to serve GPT-4o at scale — freeing compute capacity for new model development and reducing marginal serving costs.

The TBPN Acquisition: Media as Distribution

The most unusual move in OpenAI’s recent strategy is an acquisition that barely registered in mainstream coverage: the purchase of TBPN, a tech-focused business podcast and talk show network. TBPN generated approximately $5 million in advertising revenue in 2025 and was tracking toward $30 million in 2026 before the acquisition.

On its face, this makes little sense for an AI research company. Why would the developers of GPT-5 buy a podcast network?

The strategic logic becomes clearer when viewed through the lens of distribution and narrative control. OpenAI’s public narrative has been shaped almost entirely by media coverage it cannot control — coverage that has ranged from hagiographic to deeply critical, often within the same week. At $24B in annualized revenue, the company’s media relations apparatus is vastly underpowered relative to its economic and cultural influence.

TBPN gives OpenAI a direct distribution channel to a technically-sophisticated, business-oriented audience — the precise demographic that makes enterprise procurement decisions, influences developer toolchain adoption, and shapes the professional opinion-leader class’s views on AI. A $5-30M media property is rounding error at OpenAI’s scale, but the distribution and credibility it provides is genuinely scarce.

The move draws comparisons to Amazon’s Twitch acquisition (reach into a specific technical demographic) and Salesforce’s media investments in the 2010s. Whether it will be more than a curiosity in OpenAI’s history will depend on editorial independence — if TBPN’s content is perceived as company propaganda, it will be worth less than nothing. If it maintains independent credibility while benefiting from OpenAI’s resources, it could become a meaningful distribution asset.

The IPO Question

OpenAI is actively preparing for an IPO, though timing has not been publicly committed. An offering at the current $852B private valuation would be among the largest IPOs in history. For reference: Meta’s IPO in 2012 was at approximately $100 billion; Microsoft’s all-time high market cap is approximately $3 trillion; the OpenAI IPO valuation would sit roughly between Alibaba’s 2014 record-setting offering and Meta’s current market cap.

The structural complication is OpenAI’s non-profit heritage. The company is mid-transition from its original non-profit governance structure to a public benefit corporation structure, a process that has involved complex negotiations with the California and Delaware attorneys general. An IPO requires a clear, commercially-oriented corporate structure — and that structure must be in place before any offering can proceed.

The $852B valuation also creates a specific challenge: justifying it requires not just the current revenue trajectory but a credible story about how OpenAI maintains its market position as open-weight models improve, as Microsoft’s partnership evolves from strategic alignment to competitive tension, and as regulatory scrutiny intensifies on both sides of the Atlantic.

The Larger Moment

OpenAI’s revenue trajectory and the Q1 VC numbers more broadly tell a consistent story: AI has crossed from “venture bet” to “critical infrastructure” in the minds of the institutions that allocate capital. The $300 billion in global VC in Q1 2026 — with $242 billion flowing to AI — represents a reallocation of capital at a scale typically associated with prior generational shifts in infrastructure: electrification, telecommunications, the internet.

The difference from those prior eras is the speed of concentration. The top four venture rounds in Q1 2026 — OpenAI ($122B), Anthropic ($30B), xAI ($20B), Waymo ($16B) — accounted for 63% of all global VC. When two-thirds of the world’s venture capital flows to four companies in a single quarter, the dynamics of “startup ecosystem” have been fundamentally transformed.

Whether that concentration produces better AI for users or a winner-take-most dynamic that forecloses future challengers is the central question of the current moment in AI industry history. OpenAI’s $2B/month revenue figure suggests, for now, that users have voted with their wallets — and they voted decisively.