Evolving Infra: Agent-Adaptive LLM Serving at Scale A paper posted to arXiv on 9 November 2026 (revised 12 November 2026) by Hao Kang and co-authors presents Evolving Infra, an agent-adaptive LLM serving framework that decouples compute allocation from KV cache management so nodes can switch between prefill and decode at runtime. Tested on coding, math, and computer-use tasks with GLM-5.2 and Kimi K2.6 across up to 256 GB300 GPUs, Evolving Infra cut mean end-to-end request latency by up to 4.1x and P95 time-to-first-token by up to 14x versus Dynamo, NVIDIA's production framework. The system implementation is open-sourced at github.com/mit-han-lab/EvolvingInfra. Computer Science Distributed, Parallel, and Cluster Computing Submitted on 9 Nov 2026 v1 https://arxiv.org/abs/2611.12304v1 , last revised 12 Nov 2026 this version, v2 Title:Evolving Infra: Agent-Adaptive LLM Serving at Scale View PDF https://arxiv.org/pdf/2611.12304 HTML experimental https://arxiv.org/html/2611.12304v2 Abstract:Agentic workloads are a major source of LLM inference demand. Existing prefill--decode PD disaggregated serving systems offer limited flexibility for runtime adaptation of compute allocation and KV cache management, and rely on coarse signals that only partly capture the diverse and dynamic agent workload. These limitations lead to load imbalance between prefill and decode nodes and eviction of KV caches that agents will reuse. We present Evolving Infra, an agent-adaptive serving framework that adapts compute allocation to agent workloads and KV cache management to agent context management at runtime. We decouple compute allocation from key--value KV cache management, allowing nodes to switch efficiently between prefill and decode. Building on this design, an online planner matches the PD split to workload composition, while an agent-aware distributed KV manager guides cache eviction and transfer with signals from agent context management. Across coding, math, and computer-use tasks with GLM-5.2 and Kimi K2.6 on up to 256 GB300 GPUs, Evolving Infra reduces mean end-to-end request latency and P95 TTFT by up to 4.1x and 14x over Dynamo, NVIDIA's production framework. We open-source the system implementation of Evolving Infra at this https URL https://github.com/mit-han-lab/EvolvingInfra . Submission history From: Hao Kang view email https://arxiv.org/show-email/10ab0e7e/2611.12304 Mon, 9 Nov 2026 12:26:02 UTC 876 KB \ v1\ https://arxiv.org/abs/2611.12304v1 v2 Thu, 12 Nov 2026 08:20:30 UTC 867 KB Additional Features 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 .