OpenAI announced on Thursday that its business products—ChatGPT Enterprise, Team, and Edu—now have over 1 million paying users, a sharp increase from the 600,000 users reported in April. This surge in adoption highlights the success of CEO Sam Altman's strategy to encourage enterprise adoption of its AI-powered chatbot.

ChatGPT’s enterprise offerings continue to gain traction as businesses seek advanced language models to enhance operations, with the user base doubling in just a few months. OpenAI’s freemium model remains widely popular as well, with hundreds of millions of users engaging with the free version of ChatGPT every month.

In addition to its growing user base, OpenAI is reportedly considering higher-priced subscriptions for upcoming large language models (LLMs). According to a report from The Information, internal discussions have touched on pricing tiers as high as $2,000 per month for models like "Strawberry," a reasoning-focused LLM, and the flagship model "Orion."

Currently, the ChatGPT Plus subscription is priced at $20 per month, providing enhanced capabilities over the free version. However, as OpenAI continues to develop its more advanced models, it may seek to offer premium pricing to businesses requiring deeper research capabilities and more tailored AI solutions. The upcoming model "Strawberry" is said to involve specialized post-training techniques, designed to refine the AI’s performance in specific tasks after initial training on generalized data.

This potential shift in pricing comes amidst reports that OpenAI could soon raise additional capital from major investors like Apple and Nvidia, which would push the company's valuation above $100 billion. OpenAI, already backed by Microsoft, has been rapidly expanding its influence in the AI space, with its flagship ChatGPT application now boasting more than 200 million weekly active users—a figure that has doubled since last fall.

As OpenAI continues to refine its offerings and develop more sophisticated AI models, its growing base of paying users and ongoing discussions around higher-tier subscriptions signal the company’s intention to remain at the forefront of AI innovation, with an eye on deeper enterprise integration and long-term growth.

What is the Orion Model?

OpenAI’s "Orion" model is poised to be the company’s next major leap in AI, potentially succeeding the current GPT-4 model. It is expected to be a "flagship" large language model (LLM) with enhanced capabilities in language understanding, generation, and reasoning, as well as support for multimodal inputs like text, images, and video. While OpenAI has not provided an official launch date, reports suggest Orion could be unveiled as early as late 2024, with some sources pointing to a public demonstration by the end of the year​.

This release will follow the launch of "Strawberry," another significant project that focuses on advanced reasoning and research tasks. Strawberry is designed to improve complex problem-solving and generate synthetic data for model training, which is expected to be foundational for Orion​(

Orion’s development is being closely watched, particularly because of its potential to significantly outperform its predecessors in both performance and strategic applications, attracting attention from industries and government sectors alike

What is the Strawberry Model?

The "Strawberry" model is an upcoming large language model (LLM) that OpenAI is reportedly working on, aimed at enabling AI systems to perform more complex reasoning and deep research tasks. While specific technical details are still limited, the model is said to focus on advancing the reasoning capabilities of AI, making it more suited for tasks that require nuanced understanding, critical thinking, and detailed analysis beyond basic text generation.

One key feature of "Strawberry" involves what is known as "post-training" of AI models. This process allows the model to be fine-tuned or adapted after its initial training on generalized data. Post-training can help optimize the AI for specific applications or industries, making the model more specialized and effective for particular tasks, such as conducting in-depth research or solving complex problems that require domain-specific knowledge.

In essence, "Strawberry" represents a new step in tailoring AI models for more specialized uses, going beyond what the current versions of ChatGPT or other LLMs can do in general-purpose settings. This could be particularly useful for enterprise clients or researchers looking for advanced AI tools that can handle more sophisticated, critical-thinking tasks.

While the pricing for "Strawberry" has yet to be confirmed, there have been discussions of premium subscriptions reaching as high as $2,000 per month, signaling that this model will be positioned as a high-end, specialized tool for those with advanced needs in AI-driven reasoning and problem-solving.