Home · Intelligence · OpenAI Revenue
INTELLIGENCE BRIEFING

OpenAI Revenue

OpenAI's revenue trajectory, unit economics, and financial performance — pieced together from trial disclosures, press reports, and investor materials.

Updated May 13, 2026 ~30K/mo 4 key metrics 7 related dispatches
ARR (2025)
$12.7B
Growth Rate
200%+ YoY
ChatGPT Plus
$20/mo
Enterprise
Fastest segment
Executive Summary

Executive Summary

OpenAI reached $12.7 billion in annualized recurring revenue by Q4 2025, making it the fastest-growing enterprise software company ever measured. Revenue tripled year-over-year from roughly $3.7 billion in 2024, driven by explosive ChatGPT subscription growth, rapid enterprise adoption, and API consumption that exceeded internal forecasts. Despite this topline growth, the company operates at a significant net loss — burning an estimated $5–7 billion annually on compute infrastructure, talent, and research — raising fundamental questions about unit economics at scale.

Current State

OpenAI’s financial picture has come into sharper focus through an unusual combination of sources: the Musk v. Altman trial produced internal projections and margin data under discovery; investor presentations from the SoftBank-led $40 billion raise leaked through secondary market channels; and Microsoft’s quarterly earnings calls have disclosed Azure AI revenue figures that serve as a proxy for OpenAI consumption.

Revenue milestones by quarter (estimated):

PeriodARR (Est.)Key Driver
Q1 2024$3.4BChatGPT Plus crossing 20M subscribers
Q2 2024$4.2BGPT-4 Turbo API demand surge
Q3 2024$5.8BEnterprise tier launch
Q4 2024$7.5BGPT-4o consumer adoption
Q1 2025$9.1BAPI v2 pricing, o1 model launch
Q2 2025$10.5BPro tier scaling, government contracts
Q3 2025$11.8BEnterprise expansion, international growth
Q4 2025$12.7Bo3 model demand, holiday consumer surge

Monthly revenue run-rate exceeded $1 billion for the first time in October 2025. By comparison, it took Google Cloud 15 years to reach that milestone and AWS roughly 10 years.

Revenue mix (Q4 2025 estimates):

  • ChatGPT subscriptions (Plus at $20/mo, Pro at $200/mo): ~45% of revenue
  • OpenAI API (usage-based pricing): ~30% of revenue
  • Enterprise contracts (custom deployments, fine-tuning): ~20% of revenue
  • Other (licensing, partnerships): ~5% of revenue

The ChatGPT subscriber base is estimated at 30–35 million paid users across Plus and Pro tiers, with a conversion rate from free to paid of approximately 8–10%. The Pro tier at $200/month, launched in late 2024, has attracted an estimated 500,000–750,000 subscribers, disproportionately among professional developers, researchers, and financial analysts.

Key Dynamics

The Compute Cost Problem. OpenAI’s gross margins remain thin by software standards. Internal documents from the Musk trial revealed that cost of revenue — dominated by GPU compute from Microsoft Azure — consumed approximately 55–60% of topline revenue in 2024. While inference costs have declined with each model generation (GPT-4o is roughly 80% cheaper to run per token than GPT-4), usage growth has consistently outpaced efficiency gains. Each new model release triggers a demand surge that temporarily compresses margins before optimization catches up.

The Microsoft Revenue Share. Microsoft’s investment structure entitles it to a substantial share of OpenAI’s revenue until its $13 billion investment is recouped. The exact terms remain partially redacted in trial filings, but the structure is understood to work as follows: Microsoft receives a percentage of OpenAI revenue (estimated at 20–25%) that declines over time, plus Azure hosting fees that effectively recapture a further portion. This means OpenAI’s effective revenue retention is materially lower than headline figures suggest.

Enterprise as Margin Expander. Enterprise contracts carry significantly better unit economics than consumer subscriptions. Enterprise customers commit to annual contracts with minimum spending thresholds, reducing churn and improving revenue predictability. Enterprise average contract value has grown from approximately $150,000 in early 2024 to over $500,000 by late 2025, with several Fortune 100 deployments exceeding $10 million annually.

API Pricing Pressure. Competition from Anthropic (Claude), Google (Gemini), Meta (Llama open-weights models), and a growing cohort of open-source alternatives has forced progressive API price reductions. GPT-4o API pricing at launch was roughly 50% below GPT-4 Turbo, and subsequent models have continued this trend. OpenAI has partially offset pricing pressure through volume growth and by introducing premium-priced reasoning models (o1, o3) that command significantly higher per-token rates.

Who’s Involved

Sam Altman (CEO) drives revenue strategy and investor relations. Altman’s decision to launch the $200/month Pro tier was reportedly controversial internally but has proven accretive, demonstrating willingness to pay among power users. He personally negotiated the SoftBank round and several major enterprise contracts.

Microsoft (49% economic interest) serves as both OpenAI’s largest investor and its primary infrastructure provider. Satya Nadella’s strategy of embedding OpenAI models across Microsoft 365, GitHub Copilot, and Azure AI Services has created a distribution channel worth billions in indirect revenue. Microsoft reports Azure AI revenue as part of its Intelligent Cloud segment, which grew 33% year-over-year in Q1 FY2026.

Brad Lightcap (COO) oversees commercial operations, including the enterprise sales organization that has grown from roughly 50 people in early 2024 to over 400 by early 2026. Lightcap has built out vertical-specific sales teams targeting financial services, healthcare, legal, and government sectors.

Thrive Capital led the 2024 funding round and holds a board seat. Josh Kushner’s firm has been particularly active in facilitating enterprise introductions through its portfolio network.

SoftBank led the $40 billion round at a $300 billion valuation in early 2025, making Masayoshi Son’s Vision Fund the second-largest external investor after Microsoft. SoftBank’s investment included provisions for preferred infrastructure access and joint venture opportunities in Asia-Pacific markets.

What the Data Shows

Revenue per employee at OpenAI is extraordinarily high. With approximately 3,500 employees as of early 2026, the company generates roughly $3.6 million in ARR per employee — comparable to peak-era Instagram and ahead of most enterprise SaaS companies at similar scale.

Customer concentration risk is moderate. Microsoft-related revenue (including Azure AI consumption attributed to Microsoft 365 integrations) accounts for an estimated 15–20% of total revenue. No other single customer exceeds 3% of revenue, though the top 50 enterprise accounts collectively represent approximately 12% of total ARR.

Net revenue retention among enterprise customers exceeds 150%, indicating that existing customers are dramatically expanding their usage over time. This metric, disclosed in investor materials shared during the SoftBank round, is among the highest ever recorded for an enterprise software company.

Churn rates vary significantly by segment. ChatGPT Plus monthly churn is estimated at 4–6%, which is high by subscription software standards but consistent with consumer-oriented products. Enterprise annual churn is below 5%, and Pro tier churn is estimated at 2–3% monthly.

Geographic revenue distribution (estimated):

RegionRevenue ShareGrowth Rate
United States55%180% YoY
Europe20%250% YoY
Asia-Pacific15%350% YoY
Rest of World10%200% YoY

International growth is accelerating as OpenAI expands language support and establishes local data residency options to comply with regional regulations.

Revenue Comparison: OpenAI vs. Cloud Peers at Similar ARR

CompanyTime to $10B ARRYear AchievedPrimary Driver
OpenAI~3 years2025AI subscriptions + API
AWS~13 years2019Cloud infrastructure
Azure~11 years2020Enterprise cloud
Salesforce~18 years2021CRM SaaS
Google Cloud~15 years2023Cloud + AI

OpenAI’s trajectory is without precedent in enterprise technology. The closest comparison is TikTok’s advertising revenue ramp, which reached $10 billion in approximately four years, but TikTok operated in an established advertising market rather than creating a new product category.

Outlook

OpenAI’s internal projections, partially disclosed through trial filings, target $25–30 billion in revenue for 2026 and $80–100 billion by 2028. Achieving these targets depends on several assumptions: continued ChatGPT subscriber growth to 50+ million paid users, enterprise penetration of Fortune 500 companies exceeding 60%, and successful monetization of new product categories including autonomous agents, video generation (Sora), and voice/multimodal interfaces.

The path to profitability remains uncertain. OpenAI has indicated to investors that it expects to reach positive operating cash flow by 2027, contingent on inference cost reductions from custom silicon (its partnership with Broadcom for custom AI chips) and architectural efficiency improvements in next-generation models. However, the company’s research spending — estimated at $3–4 billion annually — shows no signs of moderating as the race toward artificial general intelligence intensifies.

The for-profit conversion (covered in our ownership briefing) will restructure how revenue flows to investors and could affect Microsoft’s revenue share arrangement. Any changes to the Microsoft relationship represent a material risk to OpenAI’s distribution advantage and infrastructure cost structure.

The competitive landscape is tightening. Anthropic reached approximately $2 billion ARR by early 2026, Google’s Gemini is bundled across 2+ billion user touchpoints, and open-weight models from Meta and Mistral are capturing price-sensitive API customers. OpenAI’s ability to maintain premium pricing depends on sustaining a model capability advantage that several competitors are actively closing.

Frequently Asked Questions

How much revenue does OpenAI actually make? OpenAI reached approximately $12.7 billion in annualized recurring revenue by Q4 2025, based on a triangulation of trial disclosures, investor materials, and Microsoft earnings commentary. Monthly revenue exceeded $1 billion for the first time in October 2025. The company does not file public financial statements, so all figures are estimates derived from multiple independent sources. The Musk v. Altman trial has produced the most detailed financial disclosures, including internal projections that show OpenAI targeting $25–30 billion in 2026 revenue.

Is OpenAI profitable? No. OpenAI operates at a significant net loss, estimated at $5–7 billion annually as of 2025. The primary cost drivers are GPU compute infrastructure (paid to Microsoft Azure), researcher compensation (senior ML researchers earn $1–3 million in total compensation), and ongoing research investment. Gross margins are estimated at 40–45%, which is low by software standards but improving as inference costs decline with each model generation. OpenAI has told investors it expects to reach positive operating cash flow by 2027, though this projection depends on substantial cost structure improvements.

Where does OpenAI’s revenue come from? Revenue is split across three main segments: ChatGPT consumer subscriptions (Plus at $20/month and Pro at $200/month) account for roughly 45% of revenue; OpenAI API usage-based billing accounts for approximately 30%; and enterprise contracts with custom deployments and fine-tuning represent about 20%. The remaining 5% comes from licensing agreements and partnerships. The enterprise segment is growing fastest in percentage terms, with average contract values exceeding $500,000 annually among large deployments.

How does OpenAI’s growth compare to other tech companies? OpenAI’s revenue trajectory is the fastest ever recorded for an enterprise technology company. It reached $10 billion in ARR roughly three years after launching ChatGPT, compared to 13 years for AWS, 11 years for Azure, and 18 years for Salesforce. The closest comparison in velocity is TikTok’s advertising revenue ramp, but OpenAI is building in an entirely new product category rather than an established market. Revenue per employee of approximately $3.6 million also ranks among the highest in technology history.

What happens to OpenAI’s revenue when the for-profit conversion completes? The for-profit conversion will restructure how revenue flows to investors and may alter the Microsoft revenue-sharing arrangement. Under the current capped-profit structure, investor returns are limited to a multiple of their investment (reportedly 100x for early investors, lower for later rounds). A full for-profit conversion would remove these caps, potentially making equity more valuable but also requiring OpenAI to compensate the nonprofit entity for the fair market value of its assets. The impact on day-to-day revenue operations is expected to be minimal, but the restructured entity may pursue an IPO, which would subject OpenAI to public financial reporting requirements for the first time.

Key Players 4 entities
Sam AltmanMicrosoftChatGPTOpenAI API
Data Sources 4 sources
Trial filings
Press reports
Investor materials
Secondary market data
Related Dispatches 7 dispatches
MAY 15
Releases VS Code 1.109 adds native CLAUDE.md support and Claude agent sessions VS Code 1.109 ships native support for CLAUDE.md files, Claude-specific slash commands, and Anthropic's Agent SDK integration. Claude agent sessions now run natively inside VS Code — not a hackathon project, but a phased roadmap from a senior engineer.
MAY 15
Releases Microsoft ships Power Platform skills plugin for Claude Code The microsoft/power-platform-skills repo gives Claude Code first billing over Copilot in the description itself. Ships identical plugin manifests for three platforms: Copilot, Claude Code, and OpenAI Codex.
MAY 15
Releases Microsoft announces WinUI agent plugin for Claude Code and Copilot A Principal Software Engineer at Microsoft built win-dev-skills with identical configs for Copilot, Claude Code, and OpenAI Codex. 64 commits co-authored by Copilot, 4 by Claude. The PR review skill uses cross-model verification — Claude orchestrators pass work to GPT-5.4 and vice versa.
MAY 15
Releases Anthropic ships 39 repo updates — Claude Agent SDK, API clients Anthropic pushed 39 repository updates across their GitHub organization, spanning the Claude Agent SDK, API client libraries, and supporting infrastructure.
MAY 15
Releases MCP ecosystem hits 5 active repos: spec, SDKs, servers, community index The Model Context Protocol ecosystem now spans five active repositories including the core specification, TypeScript and Python SDKs, reference servers, and a community index.
MAY 15
Money Moves OpenAI and Microsoft restructure partnership, cap revenue share OpenAI and Microsoft restructured their partnership in April, capping revenue share through 2030. The deal reshapes the financial relationship between the $13 billion investor and its most important portfolio company.
MAY 15
Money Moves Anthropic and SpaceXai strike $5B annual deal for Colossus I Anthropic signed a $5 billion annual compute deal with SpaceXai for access to the Colossus I datacenter, marking one of the largest AI infrastructure agreements to date.