AITechTalks7 months ago
What makes DeepSeek-V3.2 so efficient?
1 min read
DeepSeek-V3.2 costs $0.28 per million input tokens and $0.48 per million output tokens. It ranks fifth on the Artificial Analysis index and second among open weights models. The model uses a sparse mixture-of-experts architecture with 671 billion total parameters but only 37 billion active per token.
Level
Hype check
Tap to vote and see what everyone thinks.
