OpenAI is a San Francisco–based artificial intelligence company best known for creating ChatGPT and the GPT family of language models. Founded in December 2015 with a mission to develop artificial general intelligence (AGI) that “benefits all of humanity,” it has grown from a nonprofit research lab into one of the most valuable companies in the world — reaching a valuation of roughly $852 billion as of April 2026 and filing for an IPO in June 2026.

Where OpenAI came from

OpenAI was founded in December 2015 by a group that included Sam Altman, Elon Musk, Ilya Sutskever, Greg Brockman, and several other researchers and engineers. It launched as a nonprofit with $1 billion pledged in initial funding — though only a fraction of that was ultimately delivered.

The founding premise was straightforward: powerful AI should not become concentrated in the hands of a few companies or governments, and its benefits should be broadly shared. The word “open” in the name reflected an early commitment to publishing research freely, though the organization later restricted access to its most capable models as they became commercially significant.

In 2019, OpenAI restructured into a hybrid “capped-profit” model — limiting investor returns to 100 times their stake — to raise the capital needed for large-scale model training. In October 2025, it completed a further conversion to a Delaware-registered public benefit corporation (PBC), removing that cap. Microsoft, which has invested over $13 billion in the company, holds approximately 27%; the nonprofit OpenAI Foundation (the original entity) retains about 26%.

What OpenAI makes

OpenAI’s products span text, images, audio, and more:

  • ChatGPT — OpenAI’s flagship AI assistant, launched in November 2022. It reached 100 million users within two months — the fastest consumer software product in history to do so — and had grown to roughly one billion monthly active users by mid-2026. ChatGPT is available free at chatgpt.com, with paid plans starting at $20/month (Plus). The most advanced tier, Pro, costs $200/month (as of June 2026, per OpenAI’s pricing page) and provides near-unlimited access to frontier models.

  • GPT models — A series of large language models that power OpenAI’s products and its developer API. GPT-3 (2020) was a breakthrough in natural-language generation; GPT-4 (March 2023) added multimodal capability, accepting both text and images as input; GPT-4o (May 2024) unified text, audio, and image in a single model. The GPT-5 family, released in August 2025, extended reasoning and autonomous (“agentic”) capabilities further.

  • DALL-E — OpenAI’s image generation system, which creates images from text descriptions. DALL-E 3, integrated into ChatGPT, significantly improved how faithfully the model follows complex prompts and can render legible text within images.

  • Whisper — An open-source speech recognition system that transcribes and translates audio across dozens of languages.

Beyond consumer products, OpenAI offers a developer API that lets companies and engineers integrate its models into their own applications. Its models power thousands of third-party services, from coding assistants (GitHub Copilot was originally built on OpenAI’s Codex) to customer-service platforms to enterprise productivity tools.

How OpenAI is organized

After the October 2025 conversion to a public benefit corporation, OpenAI is governed by a board that includes representatives from the OpenAI Foundation, Microsoft, and other independent directors. The nonprofit Foundation retains a minority stake and a mandate to steer the company toward its original mission.

Sam Altman serves as CEO. In November 2023, the board briefly removed him from that role — an episode that lasted just five days before near-unanimous employee pressure and investor pushback led to his reinstatement. The incident illustrated the tensions that can arise between a mission-driven research mandate and the realities of running a large commercial organization.

Over 2024, several prominent co-founders and senior safety researchers departed, including Ilya Sutskever (who went on to found Safe Superintelligence Inc.) and John Schulman. Their exits prompted ongoing debate about how OpenAI balances speed of deployment against its stated commitment to AI safety.

Why OpenAI matters

OpenAI’s products — particularly ChatGPT and the GPT API — triggered a wave of AI adoption across industries. Its revenue grew from $6 billion in 2024 to over $20 billion in 2025, making it one of the fastest-growing technology companies ever.

Beyond products, OpenAI occupies a central position in ongoing debates about AI safety, the right pace of AI development, and how governments should regulate powerful AI systems. It has published influential safety research and has publicly advocated for government oversight of advanced AI — while simultaneously accelerating model capabilities, a tension that critics and analysts continue to examine closely.

In the news

OpenAI’s most recent frontier release, GPT-5.6, was previewed in June 2026 under a limited, government-supervised arrangement — an example of how its model releases now involve coordination with U.S. authorities. See our report: OpenAI Previews GPT-5.6 Under US Government Restrictions.

FAQ

Is ChatGPT free to use?
Yes — ChatGPT is free at chatgpt.com with usage limits. Paid plans unlock faster responses, higher usage limits, and access to the most capable models: Plus at $20/month and Pro at $200/month (as of June 2026, per OpenAI’s pricing page).

Who owns OpenAI?
Since October 2025, OpenAI is a public benefit corporation. Microsoft holds approximately 27%, the nonprofit OpenAI Foundation holds about 26%, and the remainder is distributed among employees and investors.

Is OpenAI the same as ChatGPT?
No — OpenAI is the company; ChatGPT is its most popular product. OpenAI also makes DALL-E, Whisper, and offers developer access to its models via an API.

What is the difference between ChatGPT and GPT?
GPT refers to the underlying language model (e.g. GPT-4, GPT-5); ChatGPT is the conversational interface built on top of it. Developers access the raw model through the API; most users interact with it through the ChatGPT interface.

Sources: OpenAI — Wikipedia, chatgpt.com/pricing