OpenAI to Launch First Open-Weight Language Model Since GPT-2

OpenAI has announced plans to release a new open-weight AI language model—its first since GPT-2—as it shifts toward greater collaboration with the AI community.

In a feedback form published Monday, OpenAI invited developers, researchers, and broader community members to share their input on the upcoming model. The form includes questions about user experience with other open models and desired features for OpenAI’s release.

“We’re excited to collaborate with developers, researchers, and the broader community to gather inputs and make this model as useful as possible,” OpenAI wrote on its site.

The company will hold feedback sessions and demo events beginning in San Francisco in the coming weeks, followed by additional sessions in Europe and the Asia-Pacific region.

Competitive Pressure from Open AI Labs

OpenAI’s move comes amid growing pressure from rivals like Meta and China’s DeepSeek, who have found success with open-weight AI models. Meta’s Llama models have surpassed 1 billion downloads, and DeepSeek has quickly gained global traction and investor interest.

Unlike OpenAI’s traditionally closed approach, these competitors allow broader experimentation and, in some cases, commercial use of their models—drawing significant attention from developers and enterprises alike.

Altman Acknowledges Shift Needed

During a recent Reddit Q&A, OpenAI CEO Sam Altman admitted the company may have been too cautious with open-sourcing in the past.

“[I personally think we need to] figure out a different open source strategy,” Altman said. “Not everyone at OpenAI shares this view, and it’s also not our current highest priority … We will produce better models [going forward], but we will maintain less of a lead than we did in previous years.”

Altman elaborated on the upcoming model in a post on X, noting it will include reasoning capabilities similar to OpenAI’s o3-mini. He also emphasized that the model will undergo rigorous evaluation under OpenAI’s preparedness framework before release, with additional oversight due to its open nature.

“We’re excited to see what developers build and how large companies and governments use it where they prefer to run a model themselves,” Altman wrote.

What This Means

This marks a turning point for OpenAI, which has historically kept its most powerful models proprietary. By returning to open-weight releases, OpenAI is responding to pressure from rivals like Meta and DeepSeek—who have found success enabling developers to modify and run models independently.

By inviting feedback early and hosting developer sessions globally, OpenAI is positioning itself to reengage with the broader AI ecosystem—where openness, reproducibility, and customization are increasingly important to researchers, startups, and governments.

For OpenAI, this may be less about giving up its edge and more about diversifying how it stays relevant in a rapidly evolving AI landscape. Releasing a strong but responsible open model allows the company to remain competitive, regain trust among open-source advocates, and stay present in emerging markets and regulatory discussions.

This release could signal the beginning of a new balance between proprietary innovation and open collaboration—one that more AI labs may soon be forced to consider.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.