OpenAI Launches Most Powerful ChatGPT Yet—With a Shocking Price Tag

ChatGPT creator OpenAI introduced the latest iteration of its flagship AI model, o1 Pro, on Thursday—along with a new ChatGPT Pro subscription tier that features an eye-popping price tag.

This subscription tier includes several new features, such as access to OpenAI o1, o1-mini, and GPT-4o, as well as advanced voice AI capabilities. However, the new offering comes with a steep price tag: $200 per month. In comparison, the previous subscription tier, ChatGPT Plus, costs $20 for a monthly subscription.

“A lot of people, power users of ChatGPT at this point—they really use it a lot, and they want more compute than $20 a month can buy,” OpenAI CEO Sam Altman said in a video presentation. “Pro has unlimited access to our models and also things like advanced voice mode.”

While Altman claimed that o1 is the smartest AI model in the world, he added that the Pro tier enhances its capabilities even further.

“These boosts may look small, but in complex workflows where you’re really pushing the limits of these models, it’s pretty significant,” Altman said.

In an onstage interview on Wednesday with The New York Times’ Andrew Ross Sorkin at the annual DealBook Summit, Altman highlighted ChatGPT’s significant growth since its launch in November 2022.

“We have more than 300 million weekly active users,” Altman told the audience. “We have users sending more than a billion messages a day to ChatGPT, and we have a very large developer business with 1.3 million U.S. developers and many more in the world.”

From Strawberry to o1

OpenAI first unveiled o1 in September after weeks of speculation and rumors around codenames like “Strawberry” and “Orion.” The model was considered so advanced that OpenAI decided to adopt a new naming convention, moving away from GPT-3 and 4o to simply “o1.”

“My teammates and I at OpenAI are excited to finally share the full o1 model (aka Strawberry) with you all,” OpenAI research scientist Noam Brown said in a post on X (formerly known as Twitter). “It can do a little better than just counting how many r’s are in ‘strawberry.’”

OpenAI emphasized that o1 is designed to be more efficient in its reasoning, enabling faster responses.

The model also demonstrates significant improvements in reducing “major errors on difficult real-world questions by 34%” compared to its predecessor. Unlike previous models, o1 takes its time in generating responses, effectively “thinking” about the answer before giving it.

Despite the promising upgrades, some ChatGPT subscribers may hesitate to switch to the Pro tier, as the lower-cost ChatGPT Plus tier includes ChatGPT-4o, which already offers access to OpenAI’s image generator Dall-E3 and web browsing— a feature not yet available in the o1 Pro package.

Initial reviews by Decrypt following the launch of the o1-preview in September found it slower than the existing 4o model. However, that was when the new model was still in its infancy.

Is it Worth the Price Tag?

As OpenAI introduces its most advanced tools over the next 11 days during a “12 Days of OpenAI” holiday-themed rollout, the debate over generative AI value and accessibility may be just beginning.

Edited by Andrew Hayward and Sebastian Sinclair