# What Does Context Compression Cost an Agent? Interaction Costs Unrevealed by Task-Completion Metrics

arXiv: https://arxiv.org/abs/2608.16370 · HTML: https://mlsanity.com/paper/2608.16370/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Agent memory systems, one of 16 directions in LLM agents and tool use.
- Builds on 19 of its 29 references that are in the map.
- First posted on arXiv 17 Aug 2026.

## Placed in

- LLM agents and tool use / [Agent memory systems](https://mlsanity.com/agents/agent-memory-systems/)
- Efficient LLM inference and serving / [Efficient reasoning and agent inference](https://mlsanity.com/inference/efficient-reasoning-and-agent-inference/)

## Builds on (most cited first, 19 of 19)

- [AgentBoard: An Analytical Evaluation Board of Multi-turn LLM Agents](https://mlsanity.com/paper/2401.13178/) (2024)
- [LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Models](https://mlsanity.com/paper/2310.05736/) (2023)
- [Survey on Evaluation of LLM-based Agents](https://mlsanity.com/paper/2503.16416/) (2025)
- [ACON: Optimizing Context Compression for Long-horizon LLM Agents](https://mlsanity.com/paper/2510.00615/) (2025)
- [The Complexity Trap: Simple Observation Masking Is as Efficient as LLM Summarization for Agent Context Management](https://mlsanity.com/paper/2508.21433/) (2025)
- [CostBench: Evaluating Multi-Turn Cost-Optimal Planning and Adaptation in Dynamic Environments for LLM Tool-Use Agents](https://mlsanity.com/paper/2511.02734/) (2025)
- [Fundamental Limits of Prompt Compression: A Rate-Distortion Framework for Black-Box Language Models](https://mlsanity.com/paper/2407.15504/) (2024)
- [MEMENTO: Teaching LLMs to Manage Their Own Context](https://mlsanity.com/paper/2604.09852/) (2026)
- [RE-TRAC: REcursive TRAjectory Compression for Deep Search Agents](https://mlsanity.com/paper/2602.02486/) (2026)
- [The Verifier Tax: Horizon Dependent Safety Success Tradeoffs in Tool Using LLM Agents](https://mlsanity.com/paper/2603.19328/) (2026)
- [Are Tools All We Need? Unveiling the Tool-Use Tax in LLM Agents](https://mlsanity.com/paper/2605.00136/) (2026)
- [Beyond pass@1: A Reliability Science Framework for Long-Horizon LLM Agents](https://mlsanity.com/paper/2603.29231/) (2026)
- [Reclaim Evaluation: A Lossy Memory Is Worse Than an Empty One](https://mlsanity.com/paper/2606.25449/) (2026)
- [What to Keep, What to Forget: A Rate--Distortion View of Memory Compaction in LLMs and Agents](https://mlsanity.com/paper/2607.08032/) (2026)
- [Compress the Context, Keep the Commitments: A Formal Framework for Verifiable LLM Context Compression](https://mlsanity.com/paper/2605.17304/) (2026)
- [Control Under Compression: Reliability Frontiers for Tool-Using Agents](https://mlsanity.com/paper/2608.01056/) (2026)
- [Less Context, Better Agents: Efficient Context Engineering for Long-Horizon Tool-Using LLM Agents](https://mlsanity.com/paper/2606.10209/) (2026)
- [Token Reduction Is Not Cost Reduction](https://mlsanity.com/paper/2607.12161/) (2026)
- [Context, Reasoning, and Hierarchy: A Cost-Performance Study of Compound LLM Agent Design in an Adversarial POMDP](https://mlsanity.com/paper/2605.16205/) (2026)

## Abstract (arXiv)

Task completion is the standard metric for evaluating context compression, yet it is incomplete: compression can increase an agent's interaction cost by forcing it to reacquire dropped state while leaving completion statistically unchanged. We introduce a controlled runtime measurement protocol for reacquisition cost in a bounded-horizon tool-using agent. The agent acts in a deterministic planning environment under a fixed 24-turn horizon. We vary compression severity, compare a dropping operator with a fact-preserving operator, restore dropped state through controlled oracle interventions, and decompose tool calls into retrieval and execution. We evaluate three models across two task regimes. Retrieval calls increase in all six model-regime comparisons and account for almost all added interaction; five of six remain significant after Holm correction. At the prespecified 5x comparison point, completion changes are not significant in any cell. DeepSeek shows a significant completion drop only at 10x compression. GPT-5.5 is the clearest case: completion changes from 80% to 85% (p = 1.0) while retrieval increases from 21.0 to 63.9 calls (p = .002). Retention interventions further separate state quantity, state type, and content validity. Random selection is comparable to an offline hindsight oracle, while replacing retained D-state with semantically irrelevant content increases retrieval by 57% (p < .001) without a significant completion change. In a second environment, ALFWorld, sliding compression produces no retrieval surge, showing that the reacquisition signature is environment-dependent rather than intrinsic to shortening context. Overall, compression can impose hidden interaction costs when execution-relevant state becomes absent and must be reacquired, while completion alone may not expose those costs.