{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,16]],"date-time":"2026-07-16T22:12:34Z","timestamp":1784239954290,"version":"3.55.0"},"reference-count":59,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"funder":[{"name":"European Union\u2019s Horizon 2022 Mission Cancer Programme","award":["GA101104777 (ONCODIR)"],"award-info":[{"award-number":["GA101104777 (ONCODIR)"]}]},{"DOI":"10.13039\/100018996","name":"Hellenic Academic Libraries Link","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100018996","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Access"],"published-print":{"date-parts":[[2026]]},"DOI":"10.1109\/access.2026.3709895","type":"journal-article","created":{"date-parts":[[2026,7,6]],"date-time":"2026-07-06T19:50:52Z","timestamp":1783367452000},"page":"101759-101773","source":"Crossref","is-referenced-by-count":0,"title":["Toward Energy-Aware LLM Inference for Text2SQL Workflows Through Context-Driven Model Routing"],"prefix":"10.1109","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5242-0523","authenticated-orcid":false,"given":"Argyris","family":"Kokkinis","sequence":"first","affiliation":[{"name":"Center for Research and Technology Hellas (CERTH), Thessaloniki, Greece"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Charalampos","family":"Lagonidis","sequence":"additional","affiliation":[{"name":"Center for Research and Technology Hellas (CERTH), Thessaloniki, Greece"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Thanassis","family":"Mavropoulos","sequence":"additional","affiliation":[{"name":"Center for Research and Technology Hellas (CERTH), Thessaloniki, Greece"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2505-9178","authenticated-orcid":false,"given":"Stefanos","family":"Vrochidis","sequence":"additional","affiliation":[{"name":"Center for Research and Technology Hellas (CERTH), Thessaloniki, Greece"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-032-14055-5_14"},{"key":"ref2","article-title":"Hardware acceleration of llms: A comprehensive survey and comparison","author":"Koilia","year":"2024","journal-title":"arXiv:2409.03384"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1016\/j.mlwa.2020.100006"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1145\/3737873"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2026.3656535"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2025.3594218"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/HPEC58863.2023.10363447"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1002\/sd.70544"},{"key":"ref9","article-title":"A dynamic LLM-powered agent network for task-oriented agent collaboration","author":"Liu","year":"2023","journal-title":"arXiv:2310.02170"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2021.115461"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.2196\/56628"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/TE.2024.3467912"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/AIRC64931.2025.11077484"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/3731599.3767403"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/COMST.2025.3587225"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.2147\/JMDH.S483247"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/TPWRS.2025.3637538"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2024.3437365"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.54648\/cola2024025"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1613\/jair.1.18675"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.trustnlp-main.25"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1016\/j.ipm.2024.104052"},{"key":"ref23","article-title":"From natural language to SQL: Review of LLM-based text-to-SQL systems","author":"Mohammadjafari","year":"2024","journal-title":"arXiv:2410.01066"},{"key":"ref24","article-title":"Pen-SQL framework: Enhancing text-to-SQL on open-source large language models","author":"Chen","year":"2024","journal-title":"arXiv:2405.06674"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1145\/3318464.3389767"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1577"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1007\/978-981-95-3456-2_16"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2024.129293"},{"key":"ref29","article-title":"EllieSQL: Cost-efficient text-to-SQL with complexity-aware routing","author":"Zhu","year":"2025","journal-title":"arXiv:2503.22402"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1007\/978-981-95-1581-3_28"},{"key":"ref31","article-title":"Think2SQL: Reinforce LLM reasoning capabilities for Text2SQL","author":"Papicchio","year":"2025","journal-title":"arXiv:2504.15077"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-acl.823"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-75872-0_15"},{"key":"ref34","article-title":"CHASE-SQL: Multi-path reasoning and preference optimized candidate selection in text-to-SQL","author":"Pourreza","year":"2024","journal-title":"arXiv:2410.01943"},{"key":"ref35","first-page":"1","article-title":"AGENTIQL: An agent-inspired multi-expert framework for text-to-SQL generation","volume-title":"Proc. NeurIPS Workshop Efficient Reasoning","author":"Heidari"},{"key":"ref36","article-title":"Hybrid LLM: Cost-efficient and quality-aware query routing","author":"Ding","year":"2024","journal-title":"arXiv:2404.14618"},{"key":"ref37","article-title":"HEXGEN-FLOW: Optimizing LLM inference request scheduling for agentic text-to-SQL workflow","author":"Peng","year":"2025","journal-title":"arXiv:2505.05286"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-acl.641"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.findings-acl.37"},{"key":"ref40","article-title":"Datrics Text2SQL: A framework for natural language to SQL query generation","author":"Gladkykh","year":"2025","journal-title":"arXiv:2506.12234"},{"key":"ref41","article-title":"Eagle: Efficient training-free router for multi-LLM inference","author":"Zhao","year":"2024","journal-title":"arXiv:2409.15518"},{"key":"ref42","article-title":"RouteLLM: Learning to route LLMs with preference data","author":"Ong","year":"2024","journal-title":"arXiv:2406.18665"},{"key":"ref43","article-title":"A dynamic routing approach for sustainable language model inference","author":"Ziller","year":"2025"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1145\/3727200.3727217"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.1215"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/EEI63073.2024.10696753"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.clinicalnlp-1.62"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1007\/s40593-025-00460-2"},{"key":"ref49","article-title":"CHESS: Contextual harnessing for efficient SQL synthesis","author":"Talaei","year":"2024","journal-title":"arXiv:2405.16755"},{"key":"ref50","article-title":"Divide and prompt: Chain of thought prompting for text-to-SQL","author":"Liu","year":"2023","journal-title":"arXiv:2304.11556"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-89621-7_8"},{"key":"ref52","article-title":"Text2SQL is not enough: Unifying AI and databases with TAG","author":"Biswal","year":"2024","journal-title":"arXiv:2408.14717"},{"key":"ref53","article-title":"Beyond SELECT: A comprehensive taxonomy-guided benchmark for real-world text-to-SQL translation","author":"Wang","year":"2025","journal-title":"arXiv:2511.13590"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1109\/access.2025.3632276"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1109\/LCA.2024.3406038"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v40i43.41024"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D18-1425"},{"key":"ref58","first-page":"5776","article-title":"Minilm: Deep self-attention distillation for task-agnostic compression of pre-trained transformers","volume":"33","author":"Wang","year":"2020","journal-title":"Adv. neural Inf. Process. Syst."},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1109\/SC41406.2024.00022"}],"container-title":["IEEE Access"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/6287639\/11323511\/11595566.pdf?arnumber=11595566","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,16]],"date-time":"2026-07-16T21:49:14Z","timestamp":1784238554000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11595566\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":59,"URL":"https:\/\/doi.org\/10.1109\/access.2026.3709895","relation":{},"ISSN":["2169-3536"],"issn-type":[{"value":"2169-3536","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026]]}}}