# Efficient reasoning and agent inference

Area: [Efficient LLM inference and serving](https://mlsanity.com/inference/index.md). HTML: https://mlsanity.com/inference/efficient-reasoning-and-agent-inference/

## Key facts

- Efficient reasoning and agent inference: 172 papers on arXiv since 2024 (data through 7 Oct 2026).
- 22 of its 172 mapped papers (13%) cite “R-KV: Redundancy-aware KV Cache Compression for Reasoning Models” (2025, arXiv 2505.24133).
- 20 of its 172 mapped papers (12%) cite “Dynamic Early Exit in Reasoning Models” (2025, arXiv 2504.15895).
- 20 new papers 24 Sep–7 Oct.

## Most cited here

- [R-KV: Redundancy-aware KV Cache Compression for Reasoning Models](https://mlsanity.com/paper/2505.24133/) (2025), cited by 22 papers here
- [Dynamic Early Exit in Reasoning Models](https://mlsanity.com/paper/2504.15895/) (2025), cited by 20 papers here
- [KVFlow: Efficient Prefix Caching for Accelerating LLM-Based Multi-Agent Workflows](https://mlsanity.com/paper/2507.07400/) (2025), cited by 20 papers here
- [Autellix: An Efficient Serving Engine for LLM Agents as General Programs](https://mlsanity.com/paper/2502.13965/) (2025), cited by 18 papers here
- [Continuum: Efficient and Robust Multi-Turn LLM Agent Scheduling with KV Cache Time-to-Live](https://mlsanity.com/paper/2511.02230/) (2025), cited by 15 papers here
- [DroidSpeak: KV Cache Sharing for Cross-LLM Communication and Multi-LLM Serving](https://mlsanity.com/paper/2411.02820/) (2024), cited by 11 papers here
- [Reasoning Path Compression: Compressing Generation Trajectories for Efficient LLM Reasoning](https://mlsanity.com/paper/2505.13866/) (2025), cited by 9 papers here
- [LazyEviction: Lagged KV Eviction with Attention Pattern Observation for Efficient Long Reasoning](https://mlsanity.com/paper/2506.15969/) (2025), cited by 9 papers here
- [ThinKV: Thought-Adaptive KV Cache Compression for Efficient Reasoning Models](https://mlsanity.com/paper/2510.01290/) (2025), cited by 9 papers here
- [KVCOMM: Online Cross-context KV-cache Communication for Efficient LLM-based Multi-agent Systems](https://mlsanity.com/paper/2510.12872/) (2025), cited by 9 papers here

## New (24 Sep–7 Oct)

- [Persistent Memory in Multi-Agent LLM Inference: What It Costs, What It Buys, and When You Can Tell](https://mlsanity.com/paper/2610.07782/) (2026-10-06)
- [Can Agent Harnesses and Inference Engines Hear Each Other? The HEAR Protocol for Agentic LLM Serving](https://mlsanity.com/paper/2610.06597/) (2026-10-05)
- [The Optimization Landscape of Learning Compacted Context Models](https://mlsanity.com/paper/2610.05885/) (2026-10-05)
- [Request Order Matters: Cache-History Sensitivity in Selective KV-Cache Reuse for Rolling Agents](https://mlsanity.com/paper/2610.05833/) (2026-10-05)
- [Spend Bytes on Breadth: Precision-Count Trade-offs for Decode-Time KV Compression in Long Chain-of-Thought Reasoning](https://mlsanity.com/paper/2610.05685/) (2026-10-05)
- [AvoKV-E: Payload-Aware KV Cache Eviction for Long Reasoning](https://mlsanity.com/paper/2610.03007/) (2026-10-02)
- [Persistent Context Graphs for Efficient Memory Compaction in LLM Agents](https://mlsanity.com/paper/2609.40118/) (2026-09-30)
- [Taming Speculative Search for Test-Time Scaling in LLM Serving](https://mlsanity.com/paper/2609.39334/) (2026-09-30)
- [SparseEngine: Sparse-First Inference Engine](https://mlsanity.com/paper/2609.39068/) (2026-09-30)
- [FOCUS: Training-Free Decision-Preserving Context Compression for LLM Agents](https://mlsanity.com/paper/2609.37590/) (2026-09-29)
- [Adapting Context Compression for Long-Horizon Agents with Counterfactual Continuations](https://mlsanity.com/paper/2609.36526/) (2026-09-29)
- [KV-streams for Efficient Compaction in Agentic Reinforcement Learning](https://mlsanity.com/paper/2609.35750/) (2026-09-28)
- [AgentPerfBench: A Benchmarking and Evaluation Suite for Inference Performance of Agentic LLMs](https://mlsanity.com/paper/2609.34683/) (2026-09-28)
- [KVCMAS: Efficient KV cache Correction for Shared Context in Multi-Agent Systems](https://mlsanity.com/paper/2609.34060/) (2026-09-28)
- [PReCache: Efficient KV Cache Sharing for Multi-LoRA Agents via Low-Rank Precomputation and Neutral Reconstruction](https://mlsanity.com/paper/2609.34054/) (2026-09-28)
- [Beyond Token Savings: A Systematic Study of Context Compression in LLM Agents](https://mlsanity.com/paper/2609.32961/) (2026-09-26)
- [Prefill-Free Cross-Family KV Cache Transfer for Heterogeneous Multi-Agent LLMs](https://mlsanity.com/paper/2609.32259/) (2026-09-26)
- [Receiver-Conditioned Latent Communication gives 94% CacheBack](https://mlsanity.com/paper/2609.32046/) (2026-09-25)
- [ActKV: Efficient LLM Agents through Action-Guided KV Cache Management](https://mlsanity.com/paper/2609.31395/) (2026-09-25)
- [When Can Agents Forget Their Reasoning? ICLR for Long-Horizon Agent Context Compression](https://mlsanity.com/paper/2609.29875/) (2026-09-24)