Feed Entities Podcasts Deep Dives Developers About

[KongchangAI]

Real-time AI knowledge base. Every claim is traceable to its source, with data freshness under 24 hours.

Powered by 240+ first-hand AI sources

Knowledge

Feed
Hot Entities
Podcasts

Developers

CLI
MCP Server
REST API
About

© 2026 KongchangAI. The freshest, most accurate AI knowledge

Unverified85% confidenceFactTime unknown

DeepSeek V4 Flash拥有284B总参数，每次激活13B，显存需求约150G

1

Sources

85%

Confidence

Medium-term (~90 days)

Relevance

5/31/2026

First Seen

Valid until: 8/29/2026

Sources

DeepSeek V4深度解析：混合注意力+流形约束+MOM优化器三大创新全解读

bilibili赋范大模型课堂

Related Entities

DeepSeek V4 Flashmodel

Related Claims

UnverifiedDeepSeek-V4 Pro版参数达1.6T，每次推理激活49B；Flash版参数284B，激活13B82% similar UnverifiedDeepSeek V4 Flash在OpenModel平台的速率限制为每分钟10次请求（10 RPM）和每分钟100K tokens（100K TPM）78% similar UnverifiedDeepSeek V4 Flash的UD-Q8_K_XL量化版本文件体积为162GB，采用MXFP4+BF16混合精度，达到完全无损77% similar VerifiedDeepSeek R1 7B版本原始FP16精度约需14GB存储，量化后仅需4.7GB77% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/13872

API

curl https://kongchang.com/api/v1/knowledge/claims/13872

MCP

get_claim(id=13872)