AI startup unicorn OpenAI is now ready to make money off its popular AI chatbot released in November.
OpenAI has launched its first subscription plan for ChatGPT, opening up the popular AI model for business use just two months after its public launch.
The new plan, called ChatGPT Plus and announced in a company blog post on Wednesday, will charge subscribers $20 for monthly use of ChatGPT’s tools. Users will get priority access to ChatGPT “during peak times,” faster responses, and “priority access to new features and improvements,” the company wrote. The subscription is only available initially in the United States and will roll out to a waitlist first.
OpenAI did not immediately respond to a request for comment on the announcement. The company also noted in its post that a waitlist is now available for an upcoming ChatGPT API.
The move comes just weeks after OpenAI president and chairman Greg Brockman posted to the company’s official Discord server asking users for input on ways to “monetize ChatGPT.” Some users (including this author) reported seeing an offer for a $42 per month pro tier in January, which promised early and priority access to new features and better reliability should the service, which has suffered repeated outages due to high demand in past weeks, again hit capacity.
After OpenAI launched ChatGPT in late November, more than one million users overwhelmed the tool’s site within five days, OpenAI has said. Since then, product releases and other announcements have flowed fast out of the San Francisco-based company led by CEO Sam Altman. In mid-January, the company announced a multi-year, multi-billion dollar partnership with Microsoft that reportedly reached a $10 billion commitment at a $29 billion valuation. Microsoft has since announced a slew of current and planned product integrations for its AI tools that experts see as posing a near-term challenge to Google’s popularity in products like Docs.
Earlier this week, OpenAI announced another product development, a “classifier” that it said could distinguish between human-written and AI-generated text. The company noted that the tool was very much a work in progress — unreliable with text shorter than 1,000 characters, and only correctly identifying AI text 26% of the time in its own test — but called on educators and other testers to contribute feedback and input. That request for cooperation came as New York City’s schools chancellor told press that the school system is currently considering rolling back an early-January ban on classroom use of ChatGPT.