🚨OpenAI o1 + Vision spotted in the wild Numerous reports of o1 with vision (and 200k token‼️) capabilities popping up. https://t.co/BdWvEQB1WG
Is OpenAI 01 Model out already for some people with image support ❓🤔🤔 https://t.co/deaJ6uFPeu
🚨 Wait, Did OpenAI roll out full O1? I noticed a new model available under slug "o1" » It can understand and reason upon images. » Stop token is 200002 for 4o and 4o-mini, not 199999 (o1-preview and o1-mini). I think it's a full o1 https://t.co/fI6qvZAJ0s
OpenAI is reportedly rolling out its new 'O series' models, specifically the o1 model. The o1 model, which is expected to be a significant advancement, supports image input and has been spotted with vision capabilities and 200k token capacity. This release aligns with previous statements from OpenAI employees about upcoming releases. The o1 model can understand and reason upon images, and it features a stop token of 200002 for the 4o and 4o-mini versions, differing from the 199999 stop token of the o1-preview and o1-mini versions.