DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
For developers working with large language models, DeepSeek-V4.1-Flash offers a more cost-effective solution for long-context tasks by significantly reducing KV cache requirements.
DeepSeek-V4.1-Flash, a 552B multimodal MoE model, introduces Causal Encoder-Decoder architecture to activate 16B parameters per token during decode but only 8B during prefill, improving cost efficiency for agentic workloads. It uses cross-layer KV cache reuse in CSA2 with FP4 KV caching to reduce global KV cache footprint to 890 bytes per token, roughly 1/4 of DeepSeek-V4-Flash's footprint. A dedicated deployment optimization called SWA Bounded Replay further reduces persistent KV cache footprint to roughly 1/8 of DeepSeek-V4-Flash's. Despite its smaller footprint, the model delivers better performance than the baseline. Model checkpoints are available at https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash.
The widespread adoption of long-horizon agents has made model workloads increasingly input-heavy. Although prior work has substantially reduced the cost of long-context computation, prefill remains computationally expensive, and large KV caches continue to strain HBM and SSD capacity and data-transfer bandwidth. Together, these compute, storage, and bandwidth demands constitute the primary bottleneck to further lowering deployment costs. To address this challenge, we introduce DeepSeek-V4.1-Flash, a multimodal Mixture-of-Experts (MoE) model with 552B backbone parameters and support for contexts of up to one million tokens. With its Causal Encoder-Decoder (CED) architecture, the model activates 16B parameters per token during decode but only 8B parameters during prefill, substantially improving cost efficiency for agentic workloads. To push the limits of KV cache compression, DeepSeek-V4.1-Flash combines cross-layer KV cache reuse in Compressed Sparse Attention 2 (CSA2) with FP4 KV caching. These designs reduce its global KV cache footprint (always in HBM) to 890 bytes per token, roughly 1/4 of the corresponding footprint of DeepSeek-V4-Flash. Further, through a dedicated deployment optimization known as SWA Bounded Replay, DeepSeek-V4.1-Flash reduces its persistent KV cache footprint (always on SSD or in host memory) to roughly 1/8 of that of DeepSeek-V4-Flash. Despite its much smaller KV cache footprint, the model delivers substantially better performance than the baseline. In addition, we streamline the DeepSeek-V4 architecture and introduce several efficient architectural extensions. We pretrain DeepSeek-V4.1-Flash on a multimodal corpus comprising 45T tokens and conduct comprehensive post-training, yielding strong performance across diverse text-based and multimodal agentic scenarios. Model checkpoints are available at https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash.
