| title | Beyond Semantic Similarity: Performance and Costs of Agentic Retrieval for Complex Tasks |
|---|---|
| authors | Reza Esfandiarpoor, Radek Osmulski, Yauhen Babakhin, Gabriel de Souza P. Moreira, Oliver Holworthy, Jie He, Ronay Ak, Jiarui Cai, Ryan Chesler, Bo Liu, Even Oldridge |
| date | 2026-10-05 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2610.05750 |
| affiliation | NVIDIA |
| upvotes | 17 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | Modern information systems, including many agentic workflows, use dense retrieval to explore large amounts of unstructured data. However, dense retrieval relies on surface-level semantic similarity, which is insufficient for increasingly complex search applications. Here, we investigate agentic retrieval that combines the reasoning capabilities of Large Language Models (LLMs) with the efficient corpus exploration of retrievers in a ReAct agentic loop to solve complex retrieval tasks. In our experiments, we show that agentic retrieval is more effective than standard retrieval, improving nDCG@10 by 8.7 points using the same embedding model. Moreover, while specialized retrieval methods struggle on out-of-domain tasks, agentic retrieval is highly generalizable: the same pipeline achieves competitive results on both the ViDoRe v3 and BRIGHT leaderboards. However, this improvement comes at a cost. On average, agentic retrieval takes 107.4 seconds, compared to 0.67 seconds for standard retrieval, and consumes 764.1K input and 5.8K output tokens per query. In short, our study demonstrates the effectiveness of agentic retrieval in modern data systems and motivates future work on more cost-efficient retrieval agents for large-scale deployment. |
cite
@misc{esfandiarpoor2026beyond,
title={Beyond Semantic Similarity: Performance and Costs of Agentic Retrieval for Complex Tasks},
author={Reza Esfandiarpoor and Radek Osmulski and Yauhen Babakhin and Gabriel de Souza P. Moreira and Oliver Holworthy and Jie He and Ronay Ak and Jiarui Cai and Ryan Chesler and Bo Liu and Even Oldridge},
year={2026},
eprint={2610.05750},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2610.05750}
}