Kimi Linear: An Expressive, Efficient Attention Architecture

Researchers at Moonshot AI introduced Kimi Linear, a hybrid linear attention architecture that outperforms full attention across short-context, long-context, and reinforcement learning scaling regimes. The architecture's core, Kimi Delta Attention (KDA), extends Gated DeltaNet with a finer-grained gating mechanism and achieves up to 75% KV cache reduction and 6x decoding throughput for 1M context. The team open-sourced the KDA kernel, vLLM implementations, and model checkpoints.

Computer Science Computation and Language Submitted on 30 Oct 2025 v1 https://arxiv.org/abs/2510.26692v1 , last revised 1 Nov 2025 this version, v2 Title:Kimi Linear: An Expressive, Efficient Attention Architecture View PDF /pdf/2510.26692 Abstract:We introduce Kimi Linear, a hybrid linear attention architecture that, for the first time, outperforms full attention under fair comparisons across various scenarios -- including short-context, long-context, and reinforcement learning RL scaling regimes. At its core lies Kimi Delta Attention KDA , an expressive linear attention module that extends Gated DeltaNet with a finer-grained gating mechanism, enabling more effective use of limited finite-state RNN memory. Our bespoke chunkwise algorithm achieves high hardware efficiency through a specialized variant of the Diagonal-Plus-Low-Rank DPLR transition matrices, which substantially reduces computation compared to the general DPLR formulation while remaining more consistent with the classical delta rule. We pretrain a Kimi Linear model with 3B activated parameters and 48B total parameters, based on a layerwise hybrid of KDA and Multi-Head Latent Attention MLA . Our experiments show that with an identical training recipe, Kimi Linear outperforms full MLA with a sizeable margin across all evaluated tasks, while reducing KV cache usage by up to 75% and achieving up to 6 times decoding throughput for a 1M context. These results demonstrate that Kimi Linear can be a drop-in replacement for full attention architectures with superior performance and efficiency, including tasks with longer input and output lengths. To support further research, we open-source the KDA kernel and vLLM implementations, and release the pre-trained and instruction-tuned model checkpoints. Submission history From: Yulun Du view email /show-email/19b5afec/2510.26692 Thu, 30 Oct 2025 16:59:43 UTC 645 KB v1 /abs/2510.26692v1 v2 Sat, 1 Nov 2025 12:05:18 UTC 691 KB References & Citations Loading... Bibliographic and Citation Tools Bibliographic Explorer What is the Explorer? https://info.arxiv.org/labs/showcase.html arxiv-bibliographic-explorer Connected Papers What is Connected Papers? https://www.connectedpapers.com/about Litmaps What is Litmaps? https://www.litmaps.co/ scite Smart Citations What are Smart Citations? https://www.scite.ai/ Code, Data and Media Associated with this Article alphaXiv What is alphaXiv? https://alphaxiv.org/ CatalyzeX Code Finder for Papers What is CatalyzeX? https://www.catalyzex.com DagsHub What is DagsHub? https://dagshub.com/ Gotit.pub What is GotitPub? http://gotit.pub/faq Hugging Face What is Huggingface? https://huggingface.co/huggingface ScienceCast What is ScienceCast? https://sciencecast.org/welcome Demos Recommenders and Search Tools Influence Flower What are Influence Flowers? https://influencemap.cmlab.dev/ CORE Recommender What is CORE? https://core.ac.uk/services/recommender arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs https://info.arxiv.org/labs/index.html .