AIBase
Home
AI NEWS
AI Tools
GEO & AEO
MCP
AI Models
Join Now
EN

AI News

View More

DeepSeek V4.1 Flash Officially Released: 552B MoE New Architecture, Fully Exceeds V4 Pro

DeepSeek releases V4.1 Flash, smallest in new architecture series with native multimodal vision. New architecture boosts capability, speed, throughput, scales up. 552B MoE, asymmetric Causal-Encoder-Decoder, input 8B/output 16B activation, HBM down to 1/4, cost cut.....

15k just now
DeepSeek V4.1 Flash Officially Released: 552B MoE New Architecture, Fully Exceeds V4 Pro

Performance Soars and Prices Drop! DeepSeek V4.1 Flash Released: Native Multimodal Visual Understanding Fully Exceeds the Pro Version

DeepSeek releases DeepSeek V4.1 Flash, its smallest lightweight multimodal model. It uses a 552B-parameter MoE architecture with a novel Causal-Encoder-Decoder design, enabling native vision understanding to improve capability ceiling, inference speed, and throughput, with potential to scale to larger models.....

13.2k just now
Performance Soars and Prices Drop! DeepSeek V4.1 Flash Released: Native Multimodal Visual Understanding Fully Exceeds the Pro Version
AIBase
Empowering the future, your artificial intelligence solution think tank
English简体中文繁體中文にほんご
FirendLinks:
AI Newsletters AI ToolsMCP ServersAI NewsAI MarketingLLM LeaderboardAI Ranking
© 2026AIBase
Business CooperationSite Map