Back to Deepseek V3
DeepSeek's V3.1 671B MoE instruct model — iterative improvement of V3 with stronger reasoning and instruction-following.
164K tokensMoEMIT
No benchmark scores available yet for this model.
DeepSeek's V3.1 671B MoE instruct model — iterative improvement of V3 with stronger reasoning and instruction-following.
No benchmark scores available yet for this model.