Appreciation for GPT-OSS-20B and Inquiry on Future Vision-Language Models
I would like to express my sincere appreciation for releasing GPT-OSS-20B. Its openness and accessibility have given me great motivation, and I have been actively experimenting and building around it with genuine excitement.
At the same time, the recent availability of strong open-source vision-language models (such as Qwen3-VL-30B-A3B) has naturally drawn more of my attention toward VL-capable systems. For many practical use cases, when the language capabilities are comparable, it is difficult not to prefer a model that can also understand images.
With this in mind, I would like to kindly ask whether OpenAI is considering the release of an open vision-language model in the spirit of GPT-OSS—something closer in capability and design philosophy to GPT-4o. Even a carefully scoped or research-focused VL model would be incredibly valuable for those of us who wish to build serious, long-term systems on open infrastructure.
Thank you again for your work on GPT-OSS-20B and for considering this request.