{"id":"https://openalex.org/W7158177143","doi":"https://doi.org/10.48550/arxiv.2604.25359","title":"The Structured Output Benchmark: A Multi-Source Benchmark for Evaluating Structured Output Quality in Large Language Models","display_name":"The Structured Output Benchmark: A Multi-Source Benchmark for Evaluating Structured Output Quality in Large Language Models","publication_year":2026,"publication_date":"2026-04-28","ids":{"openalex":"https://openalex.org/W7158177143","doi":"https://doi.org/10.48550/arxiv.2604.25359"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.25359","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.25359","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.25359","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134839564","display_name":"Abhinav Kumar Singh","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Singh, Abhinav Kumar","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5093223305","display_name":"Harsha Vardhan Khurdula","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Khurdula, Harsha Vardhan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054130350","display_name":"Yoeven D Khemlani","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Khemlani, Yoeven D","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5061064231","display_name":"Vineet Agarwal","orcid":"https://orcid.org/0000-0001-6235-5577"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Agarwal, Vineet","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5134839564"],"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/T10028","display_name":"Topic Modeling","score":0.27059999108314514,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10028","display_name":"Topic Modeling","score":0.27059999108314514,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10181","display_name":"Natural Language Processing Techniques","score":0.2198999971151352,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.06700000166893005,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/correctness","display_name":"Correctness","score":0.6360999941825867},{"id":"https://openalex.org/keywords/schema","display_name":"Schema (genetic algorithms)","score":0.6008999943733215},{"id":"https://openalex.org/keywords/json","display_name":"JSON","score":0.5501000285148621},{"id":"https://openalex.org/keywords/source-code","display_name":"Source code","score":0.5181999802589417},{"id":"https://openalex.org/keywords/parsing","display_name":"Parsing","score":0.5127000212669373},{"id":"https://openalex.org/keywords/data-modeling","display_name":"Data modeling","score":0.4097000062465668},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.40139999985694885},{"id":"https://openalex.org/keywords/data-quality","display_name":"Data quality","score":0.37779998779296875},{"id":"https://openalex.org/keywords/raw-data","display_name":"Raw data","score":0.37779998779296875}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8129000067710876},{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.6360999941825867},{"id":"https://openalex.org/C52146309","wikidata":"https://www.wikidata.org/wiki/Q7431116","display_name":"Schema (genetic algorithms)","level":2,"score":0.6008999943733215},{"id":"https://openalex.org/C2780416260","wikidata":"https://www.wikidata.org/wiki/Q2063","display_name":"JSON","level":2,"score":0.5501000285148621},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.5181999802589417},{"id":"https://openalex.org/C186644900","wikidata":"https://www.wikidata.org/wiki/Q194152","display_name":"Parsing","level":2,"score":0.5127000212669373},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4885999858379364},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.4221999943256378},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.4097000062465668},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.40139999985694885},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3831000030040741},{"id":"https://openalex.org/C24756922","wikidata":"https://www.wikidata.org/wiki/Q1757694","display_name":"Data quality","level":3,"score":0.37779998779296875},{"id":"https://openalex.org/C132964779","wikidata":"https://www.wikidata.org/wiki/Q2110223","display_name":"Raw data","level":2,"score":0.37779998779296875},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.37059998512268066},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.35370001196861267},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.35199999809265137},{"id":"https://openalex.org/C97250363","wikidata":"https://www.wikidata.org/wiki/Q235557","display_name":"File format","level":2,"score":0.34540000557899475},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.3328000009059906},{"id":"https://openalex.org/C2983685735","wikidata":"https://www.wikidata.org/wiki/Q5227355","display_name":"Data source","level":2,"score":0.3183000087738037},{"id":"https://openalex.org/C138958017","wikidata":"https://www.wikidata.org/wiki/Q190087","display_name":"Data type","level":2,"score":0.310699999332428},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.29739999771118164},{"id":"https://openalex.org/C183322885","wikidata":"https://www.wikidata.org/wiki/Q17007702","display_name":"Context model","level":3,"score":0.29190000891685486},{"id":"https://openalex.org/C2781252014","wikidata":"https://www.wikidata.org/wiki/Q1141900","display_name":"Unstructured data","level":3,"score":0.2903999984264374},{"id":"https://openalex.org/C133462117","wikidata":"https://www.wikidata.org/wiki/Q4929239","display_name":"Data collection","level":2,"score":0.2847999930381775},{"id":"https://openalex.org/C34736171","wikidata":"https://www.wikidata.org/wiki/Q918333","display_name":"Preprocessor","level":2,"score":0.27959999442100525},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.26820001006126404},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.25870001316070557},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.25220000743865967}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.25359","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.25359","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.25359","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.25359","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"are":[3],"increasingly":[4],"being":[5],"deployed":[6],"to":[7,24],"extract":[8],"structured":[9,31],"data":[10],"from":[11,88,107,111,119,143],"unstructured":[12],"and":[13,20,66,136,139,162,175,182,214,231],"semi-structured":[14],"sources:":[15],"parsing":[16],"invoices,":[17],"medical":[18],"records,":[19],"converting":[21],"PDF":[22],"documents":[23],"database":[25],"entries.":[26],"Yet":[27],"existing":[28],"benchmarks":[29],"for":[30],"output":[32],"generation":[33],"either":[34],"focus":[35],"on":[36,209,212,216],"schema":[37,156,194],"compliance":[38],"alone,":[39],"or":[40,91],"evaluate":[41,172],"value":[42],"correctness":[43],"within":[44],"a":[45,56,72,95,112,150,154,163,188],"single":[46],"source":[47,61,80,169,180],"domain.":[48],"We":[49,171,225],"introduce":[50],"SOB":[51],"(The":[52],"Structured":[53],"Output":[54],"Benchmark),":[55],"multi-source":[57],"benchmark":[58,100],"spanning":[59],"three":[60,179],"modalities:":[62],"native":[63],"text,":[64,135,210],"images,":[65,213],"audio":[67,141],"conversations.":[68],"All":[69],"models":[70,177,191],"receive":[71],"text-normalized":[73],"representation":[74],"of":[75,79],"their":[76],"context":[77,220],"regardless":[78],"modality;":[81],"this":[82],"deliberate":[83],"design":[84],"isolates":[85],"structured-output":[86],"capability":[87],"raw":[89],"vision":[90],"speech-processing":[92],"quality,":[93],"ensuring":[94],"fair,":[96],"source-agnostic":[97],"comparison.":[98],"Our":[99,185],"comprises":[101],"5,000":[102],"text":[103],"evaluation":[104,229],"records":[105,118,142],"derived":[106],"multi-hop":[108],"QA":[109],"drawn":[110],"25,091-record":[113],"full":[114],"corpus,":[115],"209":[116],"image":[117],"OCR-processed":[120],"PDFs":[121],"across":[122,178],"seven":[123,183],"document":[124],"types":[125],"including":[126],"multi-column":[127],"layouts,":[128],"dense":[129],"tables,":[130],"scanned":[131],"historical":[132],"documents,":[133],"small-print":[134],"mathematical":[137],"typesetting,":[138],"115":[140],"the":[144,158,168,197,227],"AMI":[145],"corpus.":[146],"Each":[147],"record":[148],"pairs":[149],"natural-language":[151],"question":[152],"with":[153],"JSON":[155],"that":[157],"model":[159],"must":[160],"follow":[161],"ground-truth":[164],"answer":[165],"verified":[166],"against":[167],"context.":[170],"21":[173],"frontier":[174],"open-weight":[176],"domains":[181],"metrics.":[184],"results":[186],"reveal":[187],"consistent":[189],"pattern:":[190],"achieve":[192],"near-perfect":[193],"compliance,":[195],"yet":[196],"best":[198],"Value":[199],"Accuracy,":[200],"measured":[201],"by":[202],"exact":[203],"leaf-value":[204],"match,":[205],"reaches":[206],"only":[207],"83.0%":[208],"67.2%":[211],"23.7%":[215],"audio,":[217],"where":[218],"longer":[219],"makes":[221],"extraction":[222],"substantially":[223],"harder.":[224],"release":[226],"dataset,":[228],"pipeline,":[230],"all":[232],"related":[233],"code.":[234]},"counts_by_year":[],"updated_date":"2026-04-30T06:11:10.768123","created_date":"2026-04-30T00:00:00"}