Despite the suggestion implied by the first word of its name, OpenAI is not exactly the most “open” company when it comes to sharing details about the inner workings of its business.

But this week, it pulled back the curtain on some key information suggesting its ChatGPT for Enterprise subscription tier is off to a strong start.

260 businesses and 150,000 employee users among them

Since first launching in late August 2023, ChatGPT for Enterprise has earned upwards of 260 business customers with more than 150,000 distinct users across them, according to new information provided by the company to Bloomberg, which conducted a wide-ranging interview with OpenAI’s chief operating officer Brad Lightcap.

The pricing for ChatGPT for Enterprise varies and is not openly disclosed on OpenAI’s website. Still, one user on Reddit was quoted at $60 per seat, so at 150,000 users, that’s $9 million a month in revenue or $108 million annualized.

That’s a substantial amount for any new offering four months after launch, but ultimately just a fraction of OpenAI’s reported total annualized revenue of $1.6 billion, according to The Information.

Who is paying for it?

And which companies exactly are using it? OpenAI did not provide a full list to Bloomberg, but it did reveal in a blog post announcing its new ChatGPT Team tier a selection of its customers including Block, Canva, Carlyle, The Estée Lauder Companies, PwC and Zapier.

The fact that OpenAI shared the number and even a few names, as well as the number itself, suggests a significant demand by the corporate world for OpenAI’s signature large language model (LLM)-powered chatbots.

That’s despite the fact some businesses have expressed concerns about using and deploying AI models due to hallucinations, regulatory, privacy and governance issues, as well as safeguarding their own company data.

(OpenAI is very clear that for its ChatGPT for Enterprise tier, “Enterprise data excluded from training by default.”)

A strong showing even against cheaper open-source offerings

It also suggests enterprise leaders and IT decision-makers are still willing to embrace OpenAI’s tech despite its leadership fracas late last year and despite the rise of compelling, powerful open-source offerings such as Meta’s Llama 2 and Mistral’s Mixtral.

That makes sense given those open source offerings tend to require more labor on the part of the user to get up and running, not to mention maintain, fine-tune, safeguard, etc. And those tend not to come with as much, or any, support plans built in.

OpenAI’s backing by Microsoft probably doesn’t hurt, either, since the latter is such a well-known, trusted, and common name in the enterprise software world, and as Microsoft is already integrating several OpenAI offerings into its own.

To put it another way: if enterprises have a choice between setting up a cheaper, yet potentially more demanding and complex open-source AI model for their employees or customers to use, or buying one “out of the box” or “off-the-shelf” from OpenAI, and which offers support and troubleshooting, they may opt for the second option and consider it money well spent.

We’ve seen this before in the consumer tech space as well: it’s possible still to this day to pirate and find pirated streams of music and TV shows, but consumers looking for security and convenience will pay a premium to get that content from Apple, Netflix, Amazon, etc.

New Team tier will likely entice even more enterprise customers

It also further supports OpenAI’s move this week to launch a new, more limited subscription tier, ChatGPT Team, where the pricing is publicly listed at $30 per user per month ($25 when billed annually).

While one could argue that OpenAI may have been better served continuing to market and enlist customers for the — presumably more expensive — ChatGPT Enterprise subscription tier, there are many companies in the small-to-medium-sized business (SMB) market that may be enticed by what appears to be a cheaper, albeit more limited, offering. (ChatGPT Enterprise offers unlimited messaging to the chatbot, while ChatGPT Team offers “higher messaging caps,” for example.)