OpenAI, the company known for bringing us the popular AI powered chatbot ChatGPT recently unveiled o1-mini, its latest large language model. Part of the company’s Project Strawberry, the new model is actually a family of new large language models that consists of two versions – o1-preview and o1-mini, a smaller and more efficient version that is aimed at code generation.
Previously limited to ChatGPT Plus, Team, Education and Enterprise edition users, OpenAI is now rolling out o1-mini for ChatGPT users who are on the free tier. The Sam Altman led company claims its new AI model is capable of reasoning and can solve complex problems faster than a human can.
While previous ChatGPT models GPT-4 were trained to offer answers by detecting patterns in the training data, the new model is based on a system of rewards and penalties. The company says o1 was also taught to process user queries by breaking them down and analyzing them one step at a time, similar to how humans do when approaching a problem.
At the moment, the new model cannot browse the internet or generate images from scratch, but OpenAI claims it is more accurate than its other large language models.
How to use o1-mini on free
ChatGPT account?
– To use o1-mini, users will have to open ChatGPT from their desktop and initiate a conversation.
– Once done, click on the option named “ChatGPT Auto” at the top of the window and under “Alpha Models”, choose “o1-mini”.
– When done, you will see that the ChatGPT Auto text has been replaced with ChatGPT Alpha (sign).
OpenAI seems to be gradually rolling out o1-mini access for free ChatGPT users, so it might take a while before the new model is available on your account.