DeepSeek Releases V4.1-Flash: New Causal Encoder-Decoder Architecture with Native Vision

DeepSeek 发布 V4.1-Flash：新 Causal Encoder-Decoder 架构，带原生视觉理解

DeepSeek V4.1-Flash is the smallest model in the new architecture family: a 552B MoE with 8B active params for input and 16B for output. It uses a Causal Encoder-Decoder design with native vision. KV cache drops to 1/4 of HBM and 1/8 of SSD storage vs the previous generation, and API pricing is lower. The post doesn't disclose exact pricing or vision benchmarks.

Why it matters: DeepSeek ships a new architecture — not a V4 refresh but a Causal Encoder-Decoder with native vision and dramatically reduced KV cache. The 552B total / 8B+16B active MoE config directly impacts deployment economics. Domestic Chinese flagship model release triggers the positiv...