DeepSeek
We present DeepSeek-V3, a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token.
Visit official pageModality
Text
License
Open Weights
Released
2025-03
AIDB Score
90
DeepSeek-V3 (Mar 2025) is built by DeepSeek and indexed in AIDB, the searchable database of every major AI model. Compare it against alternatives, find similar models, or browse other AI tools used in General.