Model Releases

DeepSeek AI releases DeepSeek-V4.1-Flash: 1M context, FP4 KV cache and cross-layer attention reuse

Original title:DeepSeek AI Released DeepSeek-V4.1-Flash with 1M Context, FP4 KV Cache, and Cross-Layer Attention Reuse

AI summary

DeepSeek AI released the multimodal MoE model DeepSeek-V4.1-Flash, with a 552B backbone plus 196B Engram parameters, a 1M context window, 8B activated parameters for prefill and 16B for decode, and a global KV cache reduced to 890 bytes per token, about 1/4 of DeepSeek-V4-Flash and 1/437 of DeepSeek-V1.

Key Takeaways

  • DeepSeek AI released the multimodal MoE model DeepSeek-V4.
  • 1-Flash,with a 552B backbone plus 196B Engram parameters,a 1M context window,8B activated…
  • and a global KV cache reduced to 890 bytes per token,about 1/4 of DeepSeek-V4-Flash and 1/437 of…

Quick Facts

Published
Updated
Source
MarkTechPost(RSS)
Category
Model Releases
AI score
87
Read the original Read the full article at the original source
Why it matters

文章拆解了 KV 缓存压缩和预精简架构的具体做法,读者可以对照其工程路径评估长上下文部署成本。

Related entities

Source & copyright

This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.

Original article · AIHOT entry

How to Cite

When citing this article, keep the title, date, and permalink, and credit Orange News as the source.

Orange News: DeepSeek AI releases DeepSeek-V4.1-Flash: 1M context, FP4 KV cache and cross-layer attention reuse. September 10, 2026. https://new.nat6.net/en/news/cmtv7kmzk02vdrok9y0y8njiv/

View all →