With an update, OpenAI is taking its “reasoning” model series out of preview and introducing a new, more costly tier for its popular chatbot ChatGPT.
The business is launching the complete version of its O1 model, which will replace the O1-preview, which was first made available as a limited preview in September under the code name “Strawberry.” Users of ChatGPT Plus and Team will be able to access the new model today, and users of Enterprise and Edu will be able to do so beginning next week.
Additionally, the business is launching ChatGPT Pro, a new $200/month subscription tier that grants users unrestricted access to Advanced Voice mode, GPT-4o, and OpenAI o1. Additionally, it comes with a version of O1 that is only available to Pro users and employs more computation to produce the optimal solution for the most challenging situations.
This version is known as O1 Pro mode. For $20 per month, the business will still provide a Plus tier that grants early access to new features, access to all of its models (except from the more potent O1 version), and other benefits.
As previously reported by The Verge, the startup intends to launch a 12-day “shipmas” period of new features, products, and demos. Among the announcements will be a new model and OpenAI’s eagerly anticipated text-to-video AI tool, Sora.
According to the business, consumers could be expecting a more accurate, powerful, and speedier model that excels at math and coding when compared to o1-preview. Additionally, it can respond to visuals with “reasoning.” Additionally, according to OpenAI, it has been trained to be more succinct, which should lead to quicker reaction times than o1-preview.
Although there is no set schedule for these modifications, OpenAI intends to integrate web surfing, file uploads, and other features into ChatGPT.
Additionally, it announced the ChatGPT Pro Grant Program, which gives 10 ChatGPT Pro funds to top medical research universities. More grants in other fields are planned.
On Thursday, OpenAI made its most sophisticated AI model to date, dubbed o1, available to paying customers. The launch marked the beginning of the company’s “12 Days of OpenAI” celebration, which consists of a dozen releases in a row to commemorate the holidays.
Upon closer inspection, the instructions are nearly worthless. The AI calculates how many inches of paint, adhesive, and sealer are needed for the job. It solely provides the measurements for the birdhouse’s front panel. For no apparent reason, it suggests cutting a piece of sandpaper to a different set of dimensions. Additionally, in a different section of the directions, it states that “the exact dimensions are as follows…” before providing no precise measurements.
OpenAI has praised o1’s reasoning skills, particularly in the areas of arithmetic and coding. In September, OpenAI released a preview version of the model along with statistics showing that it can accurately answer 78% of PhD-level science queries.
Also Read: Mobile Artificial Intelligence: Revolutionizing the Future of Smart Devices
I am Swapnil Duphare , Owner of Website : SpeedTestsWifi – I am Digital Marketer from last 6 Years and Write Blog Articles based on Internet and Speed Testing tool websites.
I am From Nashik, Maharashtra, India.