OpenAI’s Secret Strategy for Its Open Models

AI developers rejoice! After months of buildup, OpenAI yesterday finally released the first open-weight large language models it’s put out since 2019, which in AI land is forever. (ChatGPT itself was released less than three years ago.) 

The impetus for OpenAI to launch an open-weight model goes back to China’s DeepSeek, which exploded onto the scene late last year with its own open-weight model. Shortly after that release, OpenAI CEO Sam Altman seemed to acknowledge its impact, saying on Reddit that OpenAI has been “on the wrong side of history” when it came to open-source AI.

In case you’ve forgotten, “open-weight” is different from “open-source”: an open-source model lets developers freely access its weights, or the settings that control how it responds to questions, as well as its code and its training data. Open-weight models only let developers access—you guessed it—its weights.

Releasing these models as open-weight makes a lot of sense for OpenAI. First, the ChatGPT maker can get brownie points from developers and the U.S. government for providing an impressive alternative to Chinese-made open-weight models. Lots of developers I spoke to said that the performance of OpenAI’s new models matched or surpassed Chinese ones, while being significantly smaller and easier to run than some of the most popular Chinese models such as DeepSeek-V3.