deepseek-v3

1.4M 4 months ago

A strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token.

671b

5 models

deepseek-v3:latest
5da0e2d4a9e0 • 404GB • 4K context window •
Text input • 4 months ago
deepseek-v3:671b latest
5da0e2d4a9e0 • 404GB • 4K context window •
Text input • 4 months ago
deepseek-v3:671b-q4_K_M
5da0e2d4a9e0 • 404GB • 4K context window •
Text input • 4 months ago
deepseek-v3:671b-q8_0
96061c74c1a5 • 713GB • 4K context window •
Text input • 4 months ago
deepseek-v3:671b-fp16
7770bf5a5ed8 • 1.3TB • 4K context window •
Text input • 4 months ago