TLDR:
- OpenAI delays release of its first open-weight model to later in the summer.
- CEO Sam Altman says surprise research breakthroughs are the reason behind the pushback.
- The model aims to outperform top open AI competitors like DeepSeek and Mistral.
- OpenAI is under pressure to regain trust in the open-source AI community.
OpenAI has pushed back the release of its long-anticipated open-weight model, moving its launch from early summer to a later date in the season.
CEO Sam Altman made the announcement on June 10 through a post on X, revealing that the delay comes amid fresh developments from the research team. Altman described these advances as both unexpected and impressive, suggesting that the additional time will allow the company to refine the model further before public release.
we are going to take a little more time with our open-weights model, i.e. expect it later this summer but not june.
our research team did something unexpected and quite amazing and we think it will be very very worth the wait, but needs a bit longer.
— Sam Altman (@sama) June 10, 2025
While no specific release date was given, Altman’s update made it clear the model won’t be arriving in June. The delay signals that OpenAI is placing a premium on performance and impact, rather than racing to meet an earlier deadline.
Competition Heats Up in the Open-Model Arena
OpenAI’s decision to release an open-weight model comes at a time when the open-source AI landscape is growing rapidly more competitive. Rivals like Mistral and Qwen have already launched high-performing open models this year, with a strong emphasis on advanced reasoning. Mistral, in particular, introduced its “Magistral” family of models just hours before Altman’s update. These are designed to solve complex reasoning tasks, placing pressure on OpenAI to not only match but surpass them.
OpenAI’s model is expected to feature capabilities comparable to its own o-series, which powers versions of ChatGPT. Internal discussions at OpenAI have included ideas for connecting the open model to its larger, cloud-hosted AI systems for deeper query handling, although it’s unclear whether such integration will be included in the final release.
OpenAI Reputation on the Line
Beyond the technical details, the stakes are high for OpenAI’s reputation. The company has long faced criticism for its shift away from open-source principles, particularly as other labs have leaned into transparency and accessibility. Altman himself has previously admitted that OpenAI may have misjudged its approach to openness. The upcoming release is seen as an opportunity for the company to re-engage with the open-source community and reaffirm its commitment to collaborative AI development.
Success will hinge not just on how the model performs, but also on how it is received by researchers and developers eager for a powerful, transparent alternative to proprietary systems.
Growing Capacity Meets Growing Expectations
The delay also comes at a moment of significant expansion for OpenAI. Ealier this week, the company reported $10 billion in annual recurring revenue, fueled largely by its flagship product ChatGPT and associated business offerings. To meet growing computational demands, OpenAI has been diversifying its infrastructure partnerships, recently securing a deal with Google Cloud to supplement its reliance on Microsoft Azure.
These moves reflect OpenAI’s need for more computing power, not only for its commercial services but also for ambitious releases like the upcoming open-weight model. As the company scales both technically and financially, expectations around product launches have grown more intense.
