Unverified50% confidenceFactExact time

据社区讨论，DeepSeek-V4.1-Flash模型下100万token的上下文窗口占用显存仅约900MB

1

Sources

50%

Confidence

Medium-term (~90 days)

Relevance

9/11/2026

First Seen

Valid until: 12/10/2026

Sources

DeepSeek-V4.1-Flash深度解析：1M上下文仅需900MB显存的秘密

Related Entities

DeepSeek-V4.1-Flashmodel DeepSeekcompany

Related Claims

UnverifiedDeepSeek Flash模型每秒token输出可达100多73% similar Unverified经过压缩后，DeepSeek V4 Flash模型体积从接近500GB降至80.8GB71% similar Unverified据社区援引信息，DeepSeek-V4.1-Flash主干参数为552B，Engram参数为196B，safetensors权重文件总计约350GB71% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/893526

API

curl https://kongchang.com/api/v1/knowledge/claims/893526

MCP

get_claim(id=893526)