Understand any stock in simple English
Enter a ticker to learn what the company does and how it makes money.
OpenAI builds artificial intelligence models and sells access to them.
OpenAI makes artificial intelligence that people talk to. Its best-known product is ChatGPT, a chatbot you can ask questions, give tasks, or hand a messy document and get a clean answer back. It also sells access to the underlying models to software developers and businesses, so other companies can build AI into their own products. The company does not sell a physical thing, but it does sell something real: a metered connection to very capable software.
The business has two big customer groups. Regular people pay for a better version of ChatGPT, and that subscription is a simple, recurring revenue stream. Developers and companies pay based on how much text they push through the models, which is called usage-based pricing. OpenAI also struck a major deal with Microsoft, which invested billions and weaves OpenAI's models into products like Azure and Copilot. That partnership gave OpenAI money and computing power early on, but it also tied the company's fate tightly to one large partner. The hard part of this business is cost. Running these models means buying enormous amounts of computing power, mostly chips from Nvidia and servers from cloud providers. Training each new generation of model costs a fortune, and serving millions of users every day burns money constantly. That is why OpenAI has raised huge sums and why it restructured so it could attract more investment while keeping a nonprofit mission alongside a for-profit arm. Being first with ChatGPT gave it a massive lead in users and brand recognition, but competitors like Alphabet, Amazon, and Anthropic are spending heavily to catch up. Its edge is partly the quality of its models, partly the habit of its users, and partly the deep pockets of its partners.
OpenAI started in late 2015 as a non-profit research lab, founded by Sam Altman, Elon Musk, Greg Brockman and Ilya Sutskever. Its mission sounded almost impossibly big: make sure that artificial general intelligence, a machine as smart as or smarter than people at almost everything, benefits all of humanity. It began with large funding promises rather than revenue, because there was no product to sell yet, just a belief about where AI was heading.
Research is expensive, so in 2019 OpenAI built a strange new structure called a capped-profit company. Investors could earn money, but only up to a set limit, with anything beyond that flowing back to the non-profit mission. This let OpenAI raise real capital and start turning its GPT models from lab experiments into products people actually used.
Then in 2022 came the moment that changed everything. OpenAI launched ChatGPT, a simple web page where anyone could talk to a GPT model, and it became the first generative AI chat tool to go massively global. It set off a frenzy across the entire tech world, and suddenly OpenAI was not a quiet lab but the center of the conversation.
In November 2023, the board fired CEO Sam Altman, saying he had not been candid with them. The backlash was immediate: nearly every employee signed a letter demanding he come back or they would leave for Microsoft. Days later Altman returned, and several board members were replaced, a fight over who really controls OpenAI.
By 2025 OpenAI changed its shape again. The operating business became OpenAI Group PBC, a public benefit corporation legally required to weigh shareholder returns against its public mission, while the original non-profit became the OpenAI Foundation. The point was flexibility: more room to raise capital without losing the mission on paper.
That same year OpenAI launched Project Stargate, a long-term plan to pour half a trillion dollars into AI infrastructure. It announced partnerships with a long list of heavyweights, from Microsoft and SoftBank to Oracle, NVIDIA and others. OpenAI was no longer just a software company writing models; it was moving into the business of building the massive physical backbone AI runs on.
By late 2026 the mood had cooled. OpenAI cut its funding target from $30 billion to $20 billion and lowered its valuation goal, as AI and tech stocks sold off sharply and rival Anthropic prepared its own IPO. The giant raise was still giant, just smaller, a sign the easy money era was meeting reality.
The central tension is simple to state and hard to solve: the product is wildly popular, but the bills are enormous. OpenAI keeps pushing into new areas — tools that reason through multi-step problems, AI that can browse and act on your behalf, and API access that lets startups build on top of it. Whether it becomes a hugely profitable company or an expensive habit it cannot stop paying for is the question investors keep asking.
Saved answer · last researched 2 Nov 2026