OpenAI announced free GPT-5 with Deep Research technology

OpenAI announced its plans regarding the development of language models and changes in the product line. The company aims to make interaction with AI easier and clearer by eliminating the complexity of choosing between models.

It was announced by OpenAI CEO Sam Altman on his page on X.

What to expect from the development of GPT-4.5 and GPT-5

OpenAI will release GPT-4.5 (internally called Orion), which will be the last model without chain-of-thought support. The further goal is to unify the o-series and GPT-series models. The new system will be able to effectively use all available tools, adapting to different tasks.

The company will also introduce GPT-5, which will integrate technologies including o3. At the same time, a separate release of the o3 model is no longer planned. ChatGPT users will be able to communicate with GPT-5 for free without restrictions at the standard intelligence level (according to the abuse prevention policy).

With a Plus subscription, an enhanced version of GPT-5 with a higher level of intelligence will be available, and with a Pro subscription, users will receive a model that supports voice, interactive canvas, search, deep research, etc.

These changes are aimed at creating a single, intuitive system that will work without complicated model selection, providing the most effective AI experience.

As a reminder, OpenAI will prepare its own chip by 2026. By creating it, the company will be able to reduce its dependence on NVIDIA in the supply of semiconductors.

We also wrote that OpenAI has added a new feature called Deep Research to ChatGPT. It is designed to conduct deep research, allowing Artificial Intelligence to work autonomously and divide research into many stages.