OpenAI intends to “simplify’ its model offerings as it prepares for the shipping of its last non-reasoning AI model, GPT 4.5. This is a rare admission by a tech company about its products not being differentiated enough.
OpenAI’s CEO Sam Altman To clarify the product roadmap, Altman wrote to X and acknowledged that some of the recent releases had muddled things. Altman wrote
: “We want to simplify our product offerings and do a better, more transparent job of sharing our roadmap.” “We want AI to just ‘work’ for you. We realize how complex our model and product offering have become.”
Altman was particularly upset with ChatGPT’s new model picker. Users tap a dropdown to select released models. The number of models that a user is able to choose depends on the subscription level.
Altman stated that his team “hates model picker just as much as you” and will bring “magic intelligence”presumably referring a single model to power a platform. Altman said that the company intends to integrate its technology including the o series of models into GPT-5 once it is released on ChatGPT.
OpenAI released GPT-4o and o1 models over the last year. It has also launched ChatGPT subscription levels, including the $200/month ChatGPT pro.
The company has also released Operator agent, Deep Research and ChatGPT Tasks, as well as a host other new features, some of which are only available on certain models or paid levels.
Altman said that OpenAI will release GPT-4.5 as its final non-chain-of thought (CoT) model. The model was internally named Orion, but Altman did not elaborate on the new capabilities of GPT-4.5.
Altman said, “After that, we have a goal to unify the o-series and GPT series models by creating systems which can use all of our tools, know whether to think long or short and be useful for many different tasks.”
When the company releases GPT-5, it will stop shipping the reasoning o3 as a standalone model.
OpenAI understands that some users may not need the full capabilities of its newest models, Altman conceded, which is why the company created a model picker in the first place. GPT-5’s capabilities will be attuned to the subscription level to replace the dropdown, he said.
“The free tier of ChatGPT will get unlimited chat access to GPT-5 at the standard intelligence setting (!!), subject to abuse thresholds,” said Altman. “Plus subscribers will be able to run GPT-5 at a higher level of intelligence, and Pro subscribers will be able to run GPT-5 at an even higher level of intelligence. These models will incorporate voice, canvas, search, deep research and more.”
Altman did not say when GPT-4.5 or GPT-5 could be released.
Customer confusion
OpenAI’s decision to simplify the way it presents its models points to an ongoing problem: not everyone understands what is expected of each model and what makes them different. OpenAI isn’t the only one to have this problem. If I didn’t report on the industry every day, I wouldn’t be able tell you why Claude’s Sonnet is superior to Haiku.
Although this confusion is more prevalent among consumers, simplifying the products makes them much easier to manage, even for power users such as AI engineers.
You will need to have an encyclopedia that shows which models are updated and which can do certain things. It could be easier for developers to experiment if they were able to integrate the latest capabilities in one model, and only allow more complex abilities depending on usage.
Want to impress your boss? VB Daily can help. We provide you with the inside scoop on what companies do with generative AI. From regulatory shifts to practical implementations, we give you the insights you need to maximize ROI.
Read our privacy policy
Thank you for subscribing. Click here to view more VB Newsletters.
An error occured.
