DeepSeek’s multimodal model is now live, and some users are already able to try it out. So far, it’s performing pretty well. Now the question is: will this one also be open-sourced? Tomorrow might be a good time!

→ View original post on X — @jiqizhixin