OpenAI delays open-weights model release to later this summer
OpenAI just pushed back the launch of its first open model in years. CEO Sam Altman announced on X that the open-weights model won’t drop in June as planned but will arrive later this summer.
Altman teased some breakthrough work from the research team that’s "unexpected and quite amazing" — worth the wait but needing extra time. 
"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
The model aims to match or beat reasoning abilities of OpenAI’s o-series and compete with other open reasoning AI like DeepSeek’s R1.
The AI open model race is heating up. Just Tuesday, Mistral dropped its Magistral family of open reasoning models. Chinese AI lab Qwen debuted hybrid reasoning models in April — they switch between deep reasoning and quick answers.
OpenAI is also considering complex add-ons. Sources told TechCrunch the open model might connect with OpenAI’s cloud-hosted models for tougher queries — though it’s unclear if that makes the final cut.
Altman has admitted OpenAI lagged on open sourcing before. This release is urgent to rebuild trust with researchers and developers by delivering a top-tier open option.
No exact date given. Just "later this summer." The race just got longer.
