Total Cost of Agency: Exact Attribution of Memory Injection Cost in Multi-Agent LLM Workflows
Factual evidence
What the source reports
Research introduces a framework to measure hidden memory injection costs in multi-agent LLM workflows.
Inspect the evidence
- Inclusion basis
- Enterprise AI
- Publisher and source type
- arXiv cs.CL — Computation and Language · RESEARCH
- Published by source
- 4 November 2026
- Collected by OneBench
- 5 Nov 2026, 02:01 UK
Stored source excerpt
arXiv:2609.23790v1 Announce Type: cross Abstract: Every node in a multi-agent large language model (LLM) workflow retrieves context from memory and injects it into its prompt,…
Short excerpt from the collected text, not the full source. Use the source link to read it in context.
The factual summary is a OneBench synthesis, not a quotation or independent verification. Collection time is not publication time. Open the source for its full context; related reporting can share the same underlying announcement.
OneBench interpretation
Institutional assessment
So what
Multi-agent LLM architectures incur hidden input token costs from memory injections that standard observability tools fail to attribute.
Do what
Review prompt architecture and token observability tools with engineering teams building multi-agent workflows to account for memory injection costs.