OpenAI has launched a new subscription plan for ChatGPT, its AI-powered chatbot platform — and it’s very, very expensive.
Confirming leaks this morning, OpenAI announced ChatGPT Pro, a new $200-per-month subscription tier that provides unlimited access to all of OpenAI’s models, including the full version of its o1 “reasoning” model.
“We think the audience for ChatGPT Pro will be the power users of ChatGPT — those who are already pushing the models to the limits of their capabilities on tasks like math, programming, and writing,” Jason Wei, a member of OpenAI’s technical staff, said during a livestreamed press conference on Thursday.
Unlike most AI, o1 and other reasoning models attempt to check their own work as they do it. This helps them avoid some of the pitfalls that normally trip up models, with the downside being that they often take longer to arrive at solutions. O1 reasons through tasks, planning ahead and performing a series of actions that help the model tease out answers.
OpenAI released a preview of o1 in September, but this new version is, generally speaking, more performant. Compared to the preview, users can expect “a faster, more powerful, and accurate reasoning model that is even better at coding and math,” an OpenAI spokesperson told TechCrunch.
Additionally, o1 can reason about image uploads now (this wasn’t possible during the preview), and has been trained to be “more concise in its thinking” to improve response times. According to OpenAI’s internal testing, o1 reduces “major errors” on “difficult real-world questions” by 34% compared to the preview version.
Oddly enough, though, the full o1 performs worse than the preview version on a number of common benchmarks. One of those benchmarks is MLE-Bench, which measures how well AI “agents” perform at machine learning engineering.
O1 doesn’t require a ChatGPT Pro subscription. As of this afternoon, all paid ChatGPT users can access o1 through the ChatGPT model selector tool.
But ChatGPT Pro subscribers will get an ostensibly better version of o1 than users who don’t shell out as much. Called o1 pro mode, it “uses more compute for the best answers to the hardest questions,” OpenAI says.
ChatGPT Pro users can access the functionality by selecting “o1 pro mode” in the model picker and asking a question directly. Since answers will take longer to generate, ChatGPT will display a progress bar and send an in-app notification if they switch away to another conversation.
O1 pro mode may simply up the “reasoning” time the model takes before it responds with an answer. In its o1 preview announcement, OpenAI said that it aimed to experiment with o1 models that reason for hours, days, or even weeks to further boost their reasoning capabilities, and this could well be a step toward that direction.
“In evaluations from external expert testers, o1 pro mode produces more reliably accurate and comprehensive responses, especially in areas like data science, programming, and case law analysis,” an OpenAI spokesperson told TechCrunch. “Compared to both o1 and o1-preview, o1 pro mode performs better on challenging machine learning benchmarks across math, science, and coding. In particular, we saw a 75% reduction in errors for easier coding competition questions more reflective of everyday programming queries.”
O1 will soon be available in OpenAI’s API, as well, with new capabilities including function calling (i.e. the ability to use outside tools) and image analysis. OpenAI says that it plans to add support for web browsing, file uploads, and more in the months ahead.
ChatGPT Pro is easily OpenAI’s priciest plan yet — and 10x the cost of ChatGPT Plus. It’s likely to be a tough sell to all but the most devoted users, considering many people already think ChatGPT Plus is too expensive.
To sweeten the pot, ChatGPT Pro also includes unlimited access to GPT-4o and Advanced Voice Mode, ChatGPT’s human-like conversational feature.
OpenAI will also give some subscriptions away for free. The company announced a program to award 10 grants of ChatGPT Pro to medical researchers at “leading institutions,” with plans for additional grants across “various disciplines” in the future.
Price hikes for the premium ChatGPT have long been rumored. OpenAI expects to charge $44 per month for ChatGPT Plus by 2029, according to reporting by The New York Times. The company has also toyed with the idea of ultra-costly business subscriptions with additional functionality and access to models under development, per The Information. Today’s news certainly supports those reports.
The aggressive moves reflect pressure on OpenAI from investors to narrow its losses. While the company’s monthly revenue reached $300 million in August, according to The New York Times, OpenAI expects to lose roughly $5 billion this year. Expenditures like staffing, office rent, and AI training infrastructure are to blame. ChatGPT alone was at one point reportedly costing OpenAI $700,000 per day.
ChatGPT remains one of OpenAI’s biggest revenue sources. The platform has over 300 million weekly active users, around 10 million of whom are paying subscribers.
TechCrunch has an AI-focused newsletter! Sign up here to get it in your inbox every Wednesday.
