KV cache compression
KV cache compression Articles
Browse 3 articles about KV cache compression.
DeepSeek-V4.1-Flash: How 890-Byte KV Cache Compression Works
DeepSeek-V4.1-Flash compresses KV cache to 890 bytes/token via a Causal Encoder-Decoder design. Full benchmarks vs GPT-5.6, Opus 5, K3, GLM-5.3.
DeepSeek-V4.1-FlashKV cache compressionDeepSeek benchmark
How to Run DeepSeek V4.1 Flash Locally: Hardware and Setup
DeepSeek V4.1 Flash cuts KV cache needs by 4x with a 552B MoE design. Here's what hardware and setup it actually takes to self-host it.
run DeepSeek V4.1 Flash locallyDeepSeek V4.1 Flash VRAMmixture of experts local model
DeepSeek V4.1 Flash Specs: KV Cache Compression Explained
DeepSeek V4.1 Flash's model card breaks down its 552B MoE design, 1M context window, and 890-byte KV cache per token in detail.
DeepSeek V4.1 Flash specsDeepSeek architectureKV cache compression