OpenAI said to project 220 million ChatGPT subscribers by 2030

OpenAI expects 220 million people to pay for ChatGPT subscriptions by 2030, up from about 35 million as of July 2025, according to internal projections.

The company is using a strategy similar to enterprise apps like Zoom and Slack, offering a free version to a large user base—now more than 800 million—and aiming to sell enterprise subscriptions to businesses.

OpenAI forecasts that 8.5% of ChatGPT’s 2.6 billion weekly active users in 2030 will convert to paid plans, compared to about 5% today.

Current ChatGPT subscription plans range from US$20 to US$200 per month, with business-focused plans also available.

OpenAI’s revenue projections for ChatGPT subscriptions have been raised to US$87 billion for 2030, from about US$10 billion this year.

The company’s valuation recently reached US$500 billion.

🔗 Source: The Information

🧠 Food for thought

Implications, context, and why it matters.

OpenAI’s subscriber goal needs much lower costs than current API prices

A 220 million subscriber goal implies about US$395 a year per user (US$87 billion ÷ 220 million). High API spend strains profit at scale 1.
GPT-4.1 runs about US$3 per 1 million input tokens and up to US$12 for output 1. Tokens are the units of text or data used to meter and bill AI model usage. A help desk bot at 10 million tokens a month would cost about US$60 to US$120 for inference only. GPT-4o lists about US$2.50 per 1 million input tokens and US$10 for output 2.
ChatGPT Plus at US$20 per month brings US$240 a year per user. Margin stays thin if inference stays pricey, so the plan needs big cuts from chips or efficiency gains 2. Rivals like DeepSeek price input near US$0.28 per 1 million tokens versus GPT-4.1 above US$3, which pressures premium rates as OpenAI chases 220 million users 1.

Enterprises will still add third-party tools to manage ChatGPT at scale

ChatGPT Enterprise has admin controls and connectors. Many regulated firms still add third-party governance 3. Vendors can add identity plus access layers that tie ChatGPT to Single sign-on (SSO) with audit trails and role-based permissions.
Cost tools such as CloudEagle.ai, a software-as-a-service (SaaS) spend management vendor, already track SaaS usage 2. Similar platforms can watch per-user token use and enforce budget caps. Security teams can layer data loss prevention (DLP) to scan prompts and block regulated outputs.