DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...
In / out price$0.259/$0.80per 1M tokens
Context
164K tokens
Parameters
685.4BMoE256 experts · 8 per token
Modalities
→
Released
Dec 1, 2025
0 up · 0 downCompare with another model
Score history starts tracking today — check back tomorrow to see how this model moves.
Reviews (0)
Sign in to vote and write a review.
No reviews yet — be the first to review DeepSeek V3.2.