{"id":"https://openalex.org/W7225385268","doi":"https://doi.org/10.5281/zenodo.23998340","title":"Predictive Model Routing for Generative AI: Anticipating Future Request Requirements and Proactively Selecting Suitable Language Models","display_name":"Predictive Model Routing for Generative AI: Anticipating Future Request Requirements and Proactively Selecting Suitable Language Models","publication_year":2026,"publication_date":"2026-11-06","ids":{"openalex":"https://openalex.org/W7225385268","doi":"https://doi.org/10.5281/zenodo.23998340"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.23998340","doi":"https://doi.org/10.5281/zenodo.23998340","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23998340","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.23998340","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100638045","display_name":"Jincheng Zhang","orcid":"https://orcid.org/0000-0001-7332-6704","observed_orcids":["https://orcid.org/0000-0001-7332-6704","https://orcid.org/0009-0005-1860-0009"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Zhang, Jincheng","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0009-0005-1860-0009","affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5100638045"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10101","display_name":"Cloud Computing and Resource Management","score":0.7954000234603882,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10101","display_name":"Cloud Computing and Resource Management","score":0.7954000234603882,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.025699999183416367,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T14098","display_name":"Energy Efficiency in Computing","score":0.025599999353289604,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.9350000023841858},{"id":"https://openalex.org/keywords/request-classification","display_name":"request classification","score":0.6420000195503235},{"id":"https://openalex.org/keywords/model-placement","display_name":"model placement","score":0.7110000252723694},{"id":"https://openalex.org/keywords/forecast-uncertainty","display_name":"forecast uncertainty","score":0.7559999823570251},{"id":"https://openalex.org/keywords/gpu-memory-constraints","display_name":"GPU memory constraints","score":0.796999990940094},{"id":"https://openalex.org/keywords/response-quality","display_name":"response quality","score":0.5130000114440918},{"id":"https://openalex.org/keywords/inference-latency","display_name":"inference latency","score":0.5979999899864197},{"id":"https://openalex.org/keywords/cost-quality-trade-off","display_name":"cost-quality trade-off","score":0.652999997138977},{"id":"https://openalex.org/keywords/model-orchestration","display_name":"model orchestration","score":0.5339999794960022}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8259000182151794},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6097999811172485},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.5149999856948853},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4812999963760376},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.46880000829696655},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.4672999978065491},{"id":"https://openalex.org/C13606891","wikidata":"https://www.wikidata.org/wiki/Q2623243","display_name":"Conceptual model","level":2,"score":0.460099995136261},{"id":"https://openalex.org/C4679612","wikidata":"https://www.wikidata.org/wiki/Q866298","display_name":"Aggregate (composite)","level":2,"score":0.44449999928474426},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.42289999127388},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.41589999198913574},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4115000069141388},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.40450000762939453},{"id":"https://openalex.org/C74172769","wikidata":"https://www.wikidata.org/wiki/Q1446839","display_name":"Routing (electronic design automation)","level":2,"score":0.37869998812675476},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.37459999322891235},{"id":"https://openalex.org/C12186640","wikidata":"https://www.wikidata.org/wiki/Q6815743","display_name":"Memory model","level":3,"score":0.36309999227523804},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.3513000011444092},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.3422999978065491},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.33959999680519104},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3278999924659729},{"id":"https://openalex.org/C45804977","wikidata":"https://www.wikidata.org/wiki/Q7239673","display_name":"Predictive modelling","level":2,"score":0.3068000078201294},{"id":"https://openalex.org/C2776834041","wikidata":"https://www.wikidata.org/wiki/Q25346349","display_name":"Execution model","level":2,"score":0.3003999888896942},{"id":"https://openalex.org/C21338462","wikidata":"https://www.wikidata.org/wiki/Q1662581","display_name":"Information model","level":2,"score":0.2800000011920929},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.2703999876976013},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.25519999861717224},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.25119999051094055}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.23998340","doi":"https://doi.org/10.5281/zenodo.23998340","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23998340","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},{"id":"doi:10.5281/zenodo.23998341","doi":"https://doi.org/10.5281/zenodo.23998341","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23998341","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.5281/zenodo.23998340","doi":"https://doi.org/10.5281/zenodo.23998340","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23998340","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7225400169","https://openalex.org/W7225413377","https://openalex.org/W7225409378","https://openalex.org/W7225267314","https://openalex.org/W7225419693","https://openalex.org/W7225427586","https://openalex.org/W7225401292","https://openalex.org/W7225398780","https://openalex.org/W7225275234","https://openalex.org/W7225382225"],"abstract_inverted_index":{"The":[0,128,256,336],"increasing":[1],"deployment":[2],"of":[3,172,333,355,361,376],"large":[4],"language":[5],"models":[6,16,78,89],"(LLMs)":[7],"has":[8],"created":[9],"heterogeneous":[10],"inference":[11,26,53,241],"environments":[12],"in":[13,19,102,408],"which":[14,77],"multiple":[15],"differ":[17],"substantially":[18],"computational":[20,165,321],"requirements,":[21],"reasoning":[22],"capabilities,":[23,61,364],"response":[24,243,328],"quality,":[25,244,319,329],"latency,":[27,245,325],"memory":[28,250],"consumption,":[29],"and":[30,144,178,182,194,219,253,264,282,309,330,369,372,392],"operational":[31],"costs.":[32],"Conventional":[33],"model":[34,41,60,70,97,121,154,175,180,188,190,192,205,209,214,232,246,262,287,301,307,311,317,323,344,363,378,383,390,402,406],"routing":[35,122,298,345,403],"approaches":[36,50],"generally":[37],"select":[38],"an":[39,43,386],"appropriate":[40],"after":[42],"individual":[44],"user":[45],"request":[46,57,134,202,281,357],"arrives.":[47],"Although":[48],"these":[49,184],"can":[51,148],"improve":[52],"efficiency":[54],"by":[55,84],"matching":[56],"characteristics":[58,147],"with":[59,66,299],"they":[62],"remain":[63],"predominantly":[64],"reactive":[65,401],"respect":[67],"to":[68,81,151,229,277,295,389],"future":[69,85,153,170,204,270,356,382],"demand.":[71],"They":[72],"do":[73],"not":[74,340],"necessarily":[75],"anticipate":[76],"are":[79],"likely":[80],"be":[82,91,100,149],"required":[83],"requests,":[86],"when":[87],"those":[88],"will":[90],"needed,":[92],"or":[93],"whether":[94,132],"the":[95,169,275,279,285,331,353,359,365,373],"corresponding":[96,157],"resources":[98],"should":[99],"prepared":[101],"advance.":[103],"This":[104],"paper":[105],"introduces":[106],"Predictive":[107],"Model":[108],"Routing":[109],"for":[110,237],"Generative":[111],"AI":[112,410],"(PMR-GAI),":[113],"a":[114,124,398],"conceptual":[115],"framework":[116,130,257,338],"that":[117,269,284,342],"extends":[118],"conventional":[119,303],"request-level":[120],"into":[123,186],"predictive":[125,201,261,297,343,405],"decision-making":[126],"process.":[127],"proposed":[129,337],"investigates":[131,397],"historical":[133],"distributions,":[135,358],"emerging":[136],"information":[137],"trends,":[138],"temporal":[139],"usage":[140],"patterns,":[141],"application":[142],"contexts,":[143],"anticipated":[145],"task":[146],"used":[150],"forecast":[152],"requirements":[155],"before":[156],"requests":[158,173],"materialize.":[159],"Rather":[160],"than":[161],"merely":[162],"predicting":[163],"aggregate":[164],"demand,":[166],"PMR-GAI":[167],"estimates":[168],"distribution":[171],"across":[174],"capability":[176],"categories":[177],"candidate":[179],"instances,":[181],"translates":[183],"predictions":[185],"proactive":[187,213],"selection,":[189,302],"preloading,":[191],"placement,":[193],"resource":[195],"preparation":[196,254],"decisions.":[197],"Its":[198,349],"architecture":[199],"integrates":[200],"classification,":[203],"demand":[206,271,384],"forecasting,":[207],"capability-aware":[208],"matching,":[210],"cost-quality":[211],"optimization,":[212],"availability":[215],"management,":[216,394],"uncertainty-aware":[217],"routing,":[218,267,305],"continuous":[220],"feedback":[221],"adaptation.":[222],"A":[223],"unified":[224],"mathematical":[225],"formulation":[226],"is":[227,293,346],"developed":[228],"optimize":[230],"expected":[231],"selection":[233,263,318,391],"utility":[234],"while":[235],"accounting":[236],"prediction":[238,272,315],"uncertainty,":[239],"model-specific":[240],"costs,":[242],"loading":[247,324],"overhead,":[248],"GPU":[249,326],"constraints,":[251],"availability,":[252],"expenditure.":[255],"further":[258],"distinguishes":[259],"between":[260,367],"actual":[265],"request-time":[266],"recognizing":[268],"cannot":[273],"replace":[274],"need":[276],"inspect":[278],"realized":[280,370],"verify":[283],"selected":[286],"remains":[288],"appropriate.":[289],"An":[290],"experimental":[291],"methodology":[292],"established":[294],"compare":[296],"static":[300],"per-request":[304],"historical-frequency-based":[306],"preparation,":[308],"cost-aware":[310],"routing.":[312],"Evaluation":[313],"considers":[314],"accuracy,":[316],"total":[320],"expenditure,":[322],"utilization,":[327],"cost":[332],"incorrect":[334],"predictions.":[335],"does":[339],"assume":[341],"universally":[347],"beneficial.":[348],"effectiveness":[350],"depends":[351],"on":[352],"predictability":[354],"diversity":[360],"available":[362],"compatibility":[366],"predicted":[368],"workloads,":[371],"economic":[374],"value":[375],"advance":[377],"preparation.":[379],"By":[380],"introducing":[381],"as":[385],"explicit":[387],"input":[388],"infrastructure":[393],"this":[395],"research":[396],"transition":[399],"from":[400],"toward":[404],"orchestration":[407],"generative":[409],"systems.":[411]},"counts_by_year":[],"updated_date":"2026-11-08T12:12:28.789561","created_date":"2026-11-07T00:00:00"}