OpenAI is creating a new, more expensive tier for its flagship chatbot ChatGPT and removing its line of "inference" models from preview with an update. The company is releasing a full version of its o1 model (replacing the o1-preview), which was originally released as a limited preview in September (codenamed "Strawberry"). The new model will be available for ChatGPTPlus and Team users today, while Enterprise and Edu users will have access to it starting next week.
The company also launched ChatGPTPro, a new $200 monthly subscription package that includes unlimited access to OpenAIo1, GPT-4o, and advanced voice modes. It also includes an o1 version only for Pro users, which uses more calculations to provide the best answers to the hardest questions (called o1pro mode). The company will continue to offer a $20-per-month Plus package, which includes early access to new features, all of the company's models (except for the more powerful O1 version), and more.
Starting today, the startup launches a 12-day "shipmas" period of new features, products, and demos, which will include announcements of OpenAI's long-awaited text-to-video AI tool Sora and a new model.
The company says users can expect a faster, more powerful, and more accurate model than o1-preview that is superior in coding and mathematics. It can also provide "inferential" responses to images. OpenAI says it has been trained to be more concise and has faster response times than o1-preview.
OpenAI plans to add support for web browsing, file uploads, and more to ChatGPT—although there is no timetable for these changes. The company also announced a ChatGPTPro grant program that provides 10 ChatGPTPro grants to medical researchers at leading institutions, with plans for more grants across a variety of disciplines.