I don’t think it will be an end to end multimodal model, unless they’re holding that shocker for later - he says “language model” in the announcement. So basica

I don’t think it will be an end to end multimodal model, unless they’re holding that shocker for later - he says “language model” in the announcement. So basically o3 mini, not 4o multimodal.