No Multimodality yet in DeepSeek-V4. But I'll wait.

r/LocalLLaMA
Open Source AI

I hope they include it in their next v4 release.