OpenAI CEO Sam Altman recently announced that the company’s latest reasoning model, o3-mini, will be available to both free and paid ChatGPT users. This news follows Altman’s earlier confirmation that the full o3 model passed external safety testing and is slated for release as an API and ChatGPT model option. This move marks a significant shift for OpenAI, which typically reserves its newest models for paid subscribers.
This announcement comes after OpenAI first teased the o3 model in December during its “12 Days of OpenAI” event. President and Co-Founder Greg Brockman highlighted o3 as a significant breakthrough, demonstrating substantial improvement on challenging benchmarks.
Unlike traditional generative models like GPT-4 Turbo, o3 and its predecessor, o1, employ internal fact and reasoning checks before responding to user queries. This process prioritizes accuracy and reproducibility over response time, allowing these models to handle complex mathematical, scientific, and coding tasks that often challenge conventional models.
The o3-mini model is a streamlined version of the full o3 model. While smaller than GPT-4 Turbo (currently powering the paid ChatGPT experience), o3-mini is expected to offer faster response times and reduced computational demands. Providing free tier users with access to o3-mini so soon after its announcement deviates from OpenAI’s typical release strategy. Paid subscribers, while retaining access to GPT-4 Turbo, will also receive substantial access to o3-mini.
While Altman’s announcement didn’t specify the extent of paid tier access to o3-mini, this tiered access approach isn’t novel for OpenAI. Although the widespread availability of o3 is confirmed, the exact release date remains undisclosed. Currently, ChatGPT Plus subscribers pay $20 per month for access, with higher tiers like Pro ($200 per month) offering unrestricted access to models and advanced features.
