Post by Vaibhav (VB) Srivastav on X
Vaibhav (VB) Srivastav@reach_vb
XLetsss gooo! DeepSeek just released a 3B OCR model on Hugging Face 🔥
Optimised to be token efficient AND scale ~200K+ pages/day on A100-40G
Same arch as DeepSeek VL2
Use it with Transformers, vLLM and more 🤗
t.co
1.3K likes27 repliesPosted Oct 20, 2025