OpenAI will release a new “open” AI model soon

OpenAI says it plans to release a new ‘open’ AI language model in the coming months, marking the firm’s first open model since GPT-2. 

The company posted a feedback form on its website Monday to gather input from developers, researchers, and community members. This form asks devs what they would like to see in an open-weight model, and it invites them to participate in feedback sessions.

The first developer event is scheduled to take place in San Francisco in a couple weeks from now, with additional sessions planned in Europe and the Asia-Pacific regions. OpenAI hopes these gatherings will offer more insight into the needs and preferences of the community before the model’s launch and help refine its open model approach.

As pressure from rivals builds, OpenAI is eager to adapt. Chinese AI firm DeepSeek has rolled out its models in a more open manner, gaining attention from new investors. Meta’s Llama project has also gained momentum, exceeding 1 billion downloads as of March.

In a Reddit Q&A, OpenAI CEO Sam Altman acknowledged that the company has been on the “wrong side of history” regarding open sourcing its technology. While not everyone at OpenAI agrees, a different open source strategy may be necessary. Altman noted that newer models will likely be better but may not maintain as large a lead over competitors as earlier OpenAI products did.

Altman gave more details on Monday in a post on X, explaining that the open model will feature reasoning on par with the o3-mini model. He wrote that the company will run thorough checks under its preparedness framework and do even more work because the public can modify the model after release. He added that OpenAI looks forward to seeing how developers, corporations, and governments might use the model privately.

TL;DR: we are excited to release a powerful new open-weight language model with reasoning in the coming months, and we want to talk to devs about how to make it maximally useful: https://t.co/XKB4XxjREV

we are excited to make this a very, very good model!

__

we are planning to…

— Sam Altman (@sama) March 31, 2025

The feedback form also lets participants share which open models they have worked with in the past. OpenAI says the community’s perspectives will help shape its upcoming release. Altman said, “Before release, we will evaluate this model according to our preparedness framework, like we would for any other model.” 

Altman added, “And we will do extra work given that we know this model will be modified post-release. We’re excited to see what developers build and how large companies and governments use it where they prefer to run a model themselves.”

According to excerpts from an upcoming book by Wall Street Journal reporter Keach Hagey, Altman may have misled top-level colleagues about model safety procedures before his departure from OpenAI in November 2023. Though OpenAI has not addressed these claims directly, the announcement of a new open model may signal a shift toward more transparent development practices.