{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T17:12:37Z","timestamp":1780420357637,"version":"3.54.1"},"publisher-location":"New York, NY, USA","reference-count":21,"publisher":"ACM","license":[{"start":{"date-parts":[[2026,6,20]],"date-time":"2026-06-20T00:00:00Z","timestamp":1781913600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"funder":[{"name":"Austrian Science Fund (FWF)","award":["10.55776\/P36870"],"award-info":[{"award-number":["10.55776\/P36870"]}]},{"name":"Austrian Science Fund (FWF)","award":["10.55776\/PAT1668223"],"award-info":[{"award-number":["10.55776\/PAT1668223"]}]},{"name":"Vienna Science and Technology Fund (WWTF)","award":["ESR24-053"],"award-info":[{"award-number":["ESR24-053"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2026,6,21]]},"DOI":"10.1145\/3812836.3814782","type":"proceedings-article","created":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T16:54:53Z","timestamp":1780419293000},"page":"328-335","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["INAR-VL: Input-Aware Routing for Edge\u2013Cloud Vision\u2013Language Inference"],"prefix":"10.1145","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-9909-7749","authenticated-orcid":false,"given":"Ahmed","family":"\u0160abanovi\u0107","sequence":"first","affiliation":[{"name":"TU Wien, Vienna, Vienna, Austria"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-6058-8319","authenticated-orcid":false,"given":"Paul Joe","family":"Maliakel","sequence":"additional","affiliation":[{"name":"TU Wien, Vienna, Vienna, Austria"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7424-0208","authenticated-orcid":false,"given":"Ivona","family":"Brandi\u0107","sequence":"additional","affiliation":[{"name":"TU Wien, Vienna, Vienna, Austria"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2026,6,20]]},"reference":[{"key":"e_1_3_2_1_1_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2601.14921"},{"key":"e_1_3_2_1_2_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2509.23661"},{"key":"e_1_3_2_1_3_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2308.12966"},{"key":"e_1_3_2_1_4_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2410.16236"},{"key":"e_1_3_2_1_5_1","volume-title":"FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance. Trans. Mach. Learn. Res. 2024","author":"Chen Lingjiao","year":"2024","unstructured":"Lingjiao Chen, Matei Zaharia, and James Zou. 2024. FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance. Trans. Mach. Learn. Res. 2024 (2024). https:\/\/openreview.net\/forum?id=cSimKw5p6R"},{"key":"e_1_3_2_1_6_1","volume-title":"Hybrid LLM: Cost-Efficient and Quality-Aware Query Routing. In The Twelfth International Conference on Learning Representations, ICLR 2024","author":"Ding Dujian","year":"2024","unstructured":"Dujian Ding, Ankur Mallick, Chi Wang, Robert Sim, Subhabrata Mukherjee, Victor R\u00fchle, Laks V. S. Lakshmanan, and Ahmed Hassan Awadallah. 2024. Hybrid LLM: Cost-Efficient and Quality-Aware Query Routing. In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7\u201311, 2024. OpenReview.net. https:\/\/openreview.net\/forum?id=02f3mUtqnM"},{"key":"e_1_3_2_1_7_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2512.22137"},{"key":"e_1_3_2_1_8_1","doi-asserted-by":"publisher","DOI":"10.1007\/S11263-018-1116-0"},{"key":"e_1_3_2_1_9_1","volume-title":"Smoothie: Label Free Language Model Routing. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024","author":"Guha Neel","year":"2024","unstructured":"Neel Guha, Mayee F. Chen, Trevor Chow, Ishan S. Khare, and Christopher R\u00e9. 2024. Smoothie: Label Free Language Model Routing. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, NeurIPS 2024, Vancouver, BC, Canada, December 10\u201315, 2024, Amir Globersons, Lester Mackey, Danielle Belgrave, Angela Fan, Ulrich Paquet, Jakub M. Tomczak, and Cheng Zhang (Eds.). http:\/\/papers.nips.cc\/paper_files\/paper\/2024\/hash\/e6b57a990462df5afa58d64ce2709db9-Abstract-Conference.html"},{"key":"e_1_3_2_1_10_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00686"},{"key":"e_1_3_2_1_11_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICWS67624.2025.00046"},{"key":"e_1_3_2_1_12_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2405.10739"},{"key":"e_1_3_2_1_13_1","volume-title":"Generating Images with Multimodal Language Models. In Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023","author":"Koh Jing Yu","year":"2023","unstructured":"Jing Yu Koh, Daniel Fried, and Russ Salakhutdinov. 2023. Generating Images with Multimodal Language Models. In Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, NeurIPS 2023, New Orleans, LA, USA, December 10 - 16, 2023, Alice Oh, Tristan Naumann, Amir Globerson, Kate Saenko, Moritz Hardt, and Sergey Levine (Eds.). http:\/\/papers.nips.cc\/paper_files\/paper\/2023\/hash\/43a69d143273bd8215578bde887bb552-Abstract-Conference.html"},{"key":"e_1_3_2_1_14_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2406.18665"},{"key":"e_1_3_2_1_15_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2509.12492"},{"key":"e_1_3_2_1_16_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00851"},{"key":"e_1_3_2_1_17_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01202"},{"key":"e_1_3_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.1016\/J.COSREV.2025.100755"},{"key":"e_1_3_2_1_19_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2509.16995"},{"key":"e_1_3_2_1_20_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2507.15553"},{"key":"e_1_3_2_1_21_1","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.2601.17551"}],"event":{"name":"MobiSys Workshop '26: 24th Annual International Conference on Mobile Systems, Applications and Services Workshops","location":"University of Cambridge Cambridge United Kingdom","acronym":"MobiSys Workshop '26","sponsor":["SIGMOBILE ACM Special Interest Group on Mobility of Systems, Users, Data and Computing","SIGOPS ACM Special Interest Group on Operating Systems"]},"container-title":["Proceedings of the 24th Annual International Conference on Mobile Systems, Applications and Services Workshops"],"original-title":[],"deposited":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T16:57:13Z","timestamp":1780419433000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3812836.3814782"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,20]]},"references-count":21,"alternative-id":["10.1145\/3812836.3814782","10.1145\/3812836"],"URL":"https:\/\/doi.org\/10.1145\/3812836.3814782","relation":{},"subject":[],"published":{"date-parts":[[2026,6,20]]},"assertion":[{"value":"2026-06-20","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}