Another open-source model has arrived, and it’s even better than DeepSeek-V3.
The Allen Institute for AI just introduced Tülu 3 (405B) 🐫, a post-training model that is a fine-tune of Llama 3.1 405B, which outperforms DeepSeek V3.
So Deepseek is nothing but an improved GPT4 wrapper.
I'm not saying that's its bad, bust it's most likely a government funded company which burning as crazy (as GPT4 api costs are 107 times more than of deepseek v3)