DeepSeek AI Released DeepSeek-V4.1-Flash with 1M Context, FP4 KV Cache, and Cross-Layer Attention Reuse
DeepSeek released open-weight V4.1-Flash, a 552B MoE model with 1M context and FP4 KV cache, beating Opus-5 and GPT-5.6 Sol on agent benchmarks.
DeepSeek-V4.1-Flash is a multimodal Mixture-of-Experts model with a 552B-parameter backbone plus 196B Engram parameters, activating 8B parameters at prefill and 16B at decode, with a 1M-token context window. It introduces a causal encoder-decoder design, Compressed Sparse Attention 2, and FP4 (E2M1) KV cache quantization, cutting global KV cache to 890 bytes per token, about 1/4 of V4-Flash and 437x smaller than V1. Pre-training covered 45T multimodal tokens; the MIT-licensed weights ship on Hugging Face with vLLM and SGLang support. It scores 90.6 on Terminal-Bench 2.1 and 74.2 on DeepSWE v1.1, ahead of Opus-5 and GPT-5.6 Sol.
- 552B backbone MoE with 196B Engram parameters, 1M context
- 890 bytes/token KV cache, 437x smaller than DeepSeek-V1
- CSA2 shares KV across layers in Full, Reindex and Reuse modes
- Pre-trained on 45T tokens; MIT-licensed open weights
- Beats Opus-5 and GPT-5.6 Sol on agent benchmarks
- Tags
- deepseekdeepseek-v4.1-flashkv-cachelong-contextmoeopen-weightssparse-attentionspeculative-decoding
- Vendors
- DeepSeek
- Products
- vLLMSGLangTransformers
- Organizations
- Hugging Face
From the original article
4 min read at marktechpost.com
Long-horizon agents have turned LLM serving into an input-heavy workload. Repeated prefills and million-token contexts leave KV caches that strain HBM, SSD capacity, and bandwidth. DeepSeek AI built its newest release around that exact bottleneck.
By Asif Razzaq · marktechpost.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.