OpenAI CEO Sam Altman has outlined a roadmap for future AI models for the company, such as GPT-4.5 and GPT-5, in a recent post on X. Altman also spoke to user confusion about its product lineup, with OpenAI attempting to ease model selection and consolidate its product offerings.
Altman said that GPT-4.5, which had the internal codename Orion, will be the last model released by the company that won't utilize chain-of-thought reasoning. After GPT-4.5, OpenAI will merge its "o-series" and "GPT-series" models into systems capable of utilizing all the tools at their disposal.
It is worth noting that there is no release date was given for GPT-4.5 or GPT-5, but Altman suggested a timeframe of "weeks / months."
OPENAI ROADMAP UPDATE FOR GPT-4.5 and GPT-5:
— Sam Altman (@sama) February 12, 2025
We want to do a better job of sharing our intended roadmap, and a much better job simplifying our product offerings.
We want AI to “just work” for you; we realize how complicated our model and product offerings have gotten.
We hate…
The plan of the company is to integrate technologies like o3, which was initially teased in December 2024 and saw the o3-mini released in January, into the GPT-5 system. Altman also stated that o3 would no longer be pushed as an independent model.
He detailed a tiered access system for GPT-5 within ChatGPT. Free users will receive unlimited chat access at a "standard intelligence setting," while Plus subscribers will have access to GPT-5 at an enhanced intelligence level. Pro subscribers will benefit from the highest intelligence level available.
These developments come at a time when OpenAI's dominance in AI is facing growing competition. A Chinese company, DeepSeek, recently introduced its R1 reasoning model at a significantly lower cost than OpenAI's offerings.
Image source: Depositphotos
6 Comments - Add comment