#OpenAI年化经常性收入接近700亿美元 OpenAI Valuation in Perspective: $70 Billion ARR Supports $1.4 Trillion—Is It Overvalued?
OpenAI ARR (annual recurring revenue) is nearing $70 billion—up more than 70% since the beginning of Q3, with enterprise revenue doubling since July and quarterly incremental consumer revenue exceeding the full-year 2025 figure.
On the valuation side: the March financing set it at $852 billion, reports in mid-September put the target at $1.2 trillion, and the latest news is that it plans to raise $30 billion+ at a target valuation of $1.4 trillion, while Dealroom expects an IPO valuation of $2 trillion.
At $1.4 trillion / $70 billion ARR = 20x PS, it is “cheaper” than Anthropic at 30x ARR—but that does not mean OpenAI is cheap; it only shows that the entire AI sector is being priced based on “2030 revenue.” In a nutshell: the growth is real, and so is the acceleration, but the valuation is betting on whether $350 billion in revenue in 2030 can be delivered—among the AI IPO race, OpenAI is the one “running fastest and burning the most.”
I. What $70 Billion ARR Means: The Growth Is Real, and So Is the Acceleration
Three figures tell the story:
ARR rose from $40 billion in mid-August (Bloomberg) to $70 billion now—up 75% in just over a month and more than 70% since the beginning of Q3
Enterprise: Revenue has doubled since July (+100%), with 9 million paid enterprise seats
Consumer: Q3 incremental revenue exceeded the full-year 2025 increase; ChatGPT has 900 million weekly active users and more than 50 million paid subscribers
Meaning: This is not “linear growth” but “accelerating growth”—Codex programming, GPT-6 Astra subscriptions, and the enterprise market are all amplifying it. This is what gives it the confidence to ask for a $1.4 trillion valuation.
II. How the Valuation Is Calculated: 20x PS—Is It Expensive? It Depends on the Reference Point
$1.4 trillion / $70 billion ARR ≈ 20x PS (or approximately 39x PS based on projected 2026 revenue of $36 billion)
Peer comparison (AI IPO race participants):
Anthropic: ARR of approximately $6.5 billion and a target valuation of $2 trillion—approximately 30x ARR (even more expensive than OpenAI)
SpaceX: Approximately $1.75 trillion in listed market capitalization
Nvidia: Market capitalization of $4 trillion+ / revenue of $160 billion+ ≈ 25x PS
Conclusion: Within the AI sector, OpenAI’s valuation multiple is actually “middle of the pack”—the market is pricing all leading AI companies based on a “2030 story.” Whether OpenAI is cheap or expensive depends on whether you believe in its path to $350 billion in revenue in 2030 (10x its current level).
III. The AI IPO Race: Three Giants on the Same Stage—What Does It Mean for the Market?
Competitive landscape: SpaceX ($1.75T already listed), OpenAI (targeting $1.4T–2T), and Anthropic (targeting $2T)—the three companies could collectively pull hundreds of billions of dollars from the market.
Two direct impacts:
1. Liquidity drain: A concentrated wave of mega-IPOs going public could temporarily drain liquidity from the secondary market—putting pressure on high-valued tech stocks and crypto
2. Valuation anchor effect: The IPO pricing of OpenAI and Anthropic will re-anchor the valuation framework for the entire AI sector—the higher they open, the more upside imagination there will be for AI stocks in the secondary market (Nvidia, AMD, and Marvell); if they fall below their issue prices, the entire AI narrative will be hurt
IV. But OpenAI Has Two “Swords of Damocles”
Sword one: The speed of its spending. Reports say free cash flow will be negative $278 billion over the next five years—together with the compute capital expenditures promised to investors, OpenAI is essentially “buying growth with capital expenditures,” and after going public, the market will watch its cash burn every day.
Sword two: Competition + internal turmoil. Meta’s Muse is surging (the driver behind AMD breaking $1 trillion), while Google and xAI are also competing for share—OpenAI is “being scrutinized at every step, and a single mistake could shake confidence in the industry”; combined with the previous wave of executive departures (36Kr’s headline directly described it as “two sets of books on the eve of the IPO”), governance risks are significant.
V. Conclusion
OpenAI’s $70 billion ARR proves that AI commercialization is real, while its $1.4 trillion valuation proves that the market is willing to price in “AI in 2030”—but in an era of high valuations, you make money on “belief,” and you also lose money on “belief.”
OpenAI ARR (annual recurring revenue) is nearing $70 billion—up more than 70% since the beginning of Q3, with enterprise revenue doubling since July and quarterly incremental consumer revenue exceeding the full-year 2025 figure.
On the valuation side: the March financing set it at $852 billion, reports in mid-September put the target at $1.2 trillion, and the latest news is that it plans to raise $30 billion+ at a target valuation of $1.4 trillion, while Dealroom expects an IPO valuation of $2 trillion.
At $1.4 trillion / $70 billion ARR = 20x PS, it is “cheaper” than Anthropic at 30x ARR—but that does not mean OpenAI is cheap; it only shows that the entire AI sector is being priced based on “2030 revenue.” In a nutshell: the growth is real, and so is the acceleration, but the valuation is betting on whether $350 billion in revenue in 2030 can be delivered—among the AI IPO race, OpenAI is the one “running fastest and burning the most.”
I. What $70 Billion ARR Means: The Growth Is Real, and So Is the Acceleration
Three figures tell the story:
ARR rose from $40 billion in mid-August (Bloomberg) to $70 billion now—up 75% in just over a month and more than 70% since the beginning of Q3
Enterprise: Revenue has doubled since July (+100%), with 9 million paid enterprise seats
Consumer: Q3 incremental revenue exceeded the full-year 2025 increase; ChatGPT has 900 million weekly active users and more than 50 million paid subscribers
Meaning: This is not “linear growth” but “accelerating growth”—Codex programming, GPT-6 Astra subscriptions, and the enterprise market are all amplifying it. This is what gives it the confidence to ask for a $1.4 trillion valuation.
II. How the Valuation Is Calculated: 20x PS—Is It Expensive? It Depends on the Reference Point
$1.4 trillion / $70 billion ARR ≈ 20x PS (or approximately 39x PS based on projected 2026 revenue of $36 billion)
Peer comparison (AI IPO race participants):
Anthropic: ARR of approximately $6.5 billion and a target valuation of $2 trillion—approximately 30x ARR (even more expensive than OpenAI)
SpaceX: Approximately $1.75 trillion in listed market capitalization
Nvidia: Market capitalization of $4 trillion+ / revenue of $160 billion+ ≈ 25x PS
Conclusion: Within the AI sector, OpenAI’s valuation multiple is actually “middle of the pack”—the market is pricing all leading AI companies based on a “2030 story.” Whether OpenAI is cheap or expensive depends on whether you believe in its path to $350 billion in revenue in 2030 (10x its current level).
III. The AI IPO Race: Three Giants on the Same Stage—What Does It Mean for the Market?
Competitive landscape: SpaceX ($1.75T already listed), OpenAI (targeting $1.4T–2T), and Anthropic (targeting $2T)—the three companies could collectively pull hundreds of billions of dollars from the market.
Two direct impacts:
1. Liquidity drain: A concentrated wave of mega-IPOs going public could temporarily drain liquidity from the secondary market—putting pressure on high-valued tech stocks and crypto
2. Valuation anchor effect: The IPO pricing of OpenAI and Anthropic will re-anchor the valuation framework for the entire AI sector—the higher they open, the more upside imagination there will be for AI stocks in the secondary market (Nvidia, AMD, and Marvell); if they fall below their issue prices, the entire AI narrative will be hurt
IV. But OpenAI Has Two “Swords of Damocles”
Sword one: The speed of its spending. Reports say free cash flow will be negative $278 billion over the next five years—together with the compute capital expenditures promised to investors, OpenAI is essentially “buying growth with capital expenditures,” and after going public, the market will watch its cash burn every day.
Sword two: Competition + internal turmoil. Meta’s Muse is surging (the driver behind AMD breaking $1 trillion), while Google and xAI are also competing for share—OpenAI is “being scrutinized at every step, and a single mistake could shake confidence in the industry”; combined with the previous wave of executive departures (36Kr’s headline directly described it as “two sets of books on the eve of the IPO”), governance risks are significant.
V. Conclusion
OpenAI’s $70 billion ARR proves that AI commercialization is real, while its $1.4 trillion valuation proves that the market is willing to price in “AI in 2030”—but in an era of high valuations, you make money on “belief,” and you also lose money on “belief.”
