OpenAI is planning to release a new open language model in the coming months, which will be its first open model since GPT-2 in 2019. This new model will be publicly available for anyone to use, download, modify, or deploy for their own purposes.
The model is described as an "open-weight" model, which means it's not entirely open-source but still allows for more flexibility and customization compared to fully closed models like GPT-3 and GPT-4.
OpenAI CEO Sam Altman announced that the company is seeking feedback from developers and researchers to make the model as useful as possible. They're hosting developer events to gather feedback and allow attendees to play with early prototypes. These events will start in San Francisco and later expand to Europe and the Asia-Pacific region.
This release is significant, especially given the growing competition in the AI landscape with models like DeepSeek and Meta's Llama.
