{"id":"https://openalex.org/W7160533687","doi":"https://doi.org/10.48550/arxiv.2605.05185","title":"OpenSearch-VL: An Open Recipe for Frontier Multimodal Search Agents","display_name":"OpenSearch-VL: An Open Recipe for Frontier Multimodal Search Agents","publication_year":2026,"publication_date":"2026-05-06","ids":{"openalex":"https://openalex.org/W7160533687","doi":"https://doi.org/10.48550/arxiv.2605.05185"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.05185","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.05185","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":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.05185","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135568350","display_name":"Shuang Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Shuang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135593769","display_name":"Kaituo Feng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Feng, Kaituo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048946763","display_name":"Hangting Chen","orcid":"https://orcid.org/0000-0002-4085-4364"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Hangting","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135630571","display_name":"Wenxuan Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Wenxuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5010781772","display_name":"D. Z. Dai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dai, Dasen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5106748272","display_name":"Q. Y. Shou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shou, Quanxin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135628973","display_name":"Yunlong Lin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Yunlong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135608373","display_name":"Xiangyu Yue","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yue, Xiangyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135567754","display_name":"Shenghua Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Shenghua","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135566998","display_name":"Tianyu Pang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pang, Tianyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9222999811172485,"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"}},"topics":[{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9222999811172485,"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"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.010999999940395355,"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/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.009399999864399433,"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/construct","display_name":"Construct (python library)","score":0.6348999738693237},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.5414999723434448},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.5364999771118164},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.4235000014305115},{"id":"https://openalex.org/keywords/path","display_name":"Path (computing)","score":0.40610000491142273},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.39469999074935913},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.3610000014305115},{"id":"https://openalex.org/keywords/multimodal-learning","display_name":"Multimodal learning","score":0.3596999943256378}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7832000255584717},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.6348999738693237},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6078000068664551},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.5414999723434448},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.5364999771118164},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4632999897003174},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.4235000014305115},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.40610000491142273},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.39469999074935913},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.3610000014305115},{"id":"https://openalex.org/C2780660688","wikidata":"https://www.wikidata.org/wiki/Q25052564","display_name":"Multimodal learning","level":2,"score":0.3596999943256378},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3506999909877777},{"id":"https://openalex.org/C2778671685","wikidata":"https://www.wikidata.org/wiki/Q219239","display_name":"Recipe","level":2,"score":0.3506999909877777},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.3474000096321106},{"id":"https://openalex.org/C58166","wikidata":"https://www.wikidata.org/wiki/Q224821","display_name":"Fuzzy logic","level":2,"score":0.3395000100135803},{"id":"https://openalex.org/C2778571376","wikidata":"https://www.wikidata.org/wiki/Q1355821","display_name":"Frontier","level":2,"score":0.33570000529289246},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.3346000015735626},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.30959999561309814},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.2734000086784363},{"id":"https://openalex.org/C2777974031","wikidata":"https://www.wikidata.org/wiki/Q493641","display_name":"Multimodal transport","level":2,"score":0.25870001316070557},{"id":"https://openalex.org/C2777402240","wikidata":"https://www.wikidata.org/wiki/Q6783436","display_name":"Masking (illustration)","level":2,"score":0.2558000087738037}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.05185","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.05185","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":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.05185","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.05185","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":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.6285333037376404}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Deep":[0],"search":[1,30,69,221],"has":[2],"become":[3],"a":[4,60,78,124,154],"crucial":[5],"capability":[6],"for":[7,64,115,119],"frontier":[8,66],"multimodal":[9,29,67,219],"agents,":[10],"enabling":[11,141],"models":[12,202,213],"to":[13,34,38,81,143,199,214],"solve":[14],"complex":[15],"questions":[16],"through":[17,86,174],"active":[18,145],"search,":[19,131,133],"evidence":[20],"verification,":[21],"and":[22,93,101,117,138,195,212],"multi-step":[23],"reasoning.":[24],"Despite":[25],"rapid":[26],"progress,":[27],"top-tier":[28],"agents":[31,70,142],"remain":[32],"difficult":[33],"reproduce,":[35],"largely":[36],"due":[37],"the":[39],"absence":[40],"of":[41],"open":[42,216],"high-quality":[43,83],"training":[44,52,65,84,112,158],"data,":[45,210],"transparent":[46],"trajectory":[47],"synthesis":[48],"pipelines,":[49],"or":[50],"detailed":[51],"recipes.":[53],"To":[54],"this":[55,107,180],"end,":[56],"we":[57,76,109,122,152],"introduce":[58],"OpenSearch-VL,":[59],"fully":[61],"open-source":[62],"recipe":[63],"deep":[68,220],"with":[71,147,187],"agentic":[72],"reinforcement":[73],"learning.":[74],"First,":[75],"curated":[77],"dedicated":[79],"pipeline":[80],"construct":[82],"data":[85],"Wikipedia":[87],"path":[88],"sampling,":[89],"fuzzy":[90],"entity":[91],"rewriting,":[92],"source-anchor":[94],"visual":[95],"grounding,":[96],"which":[97],"jointly":[98],"reduce":[99],"shortcuts":[100],"one-step":[102],"retrieval":[103],"collapse.":[104],"Based":[105],"on":[106,179,203,218],"pipeline,":[108],"curate":[110],"two":[111],"datasets,":[113],"SearchVL-SFT-36k":[114],"SFT":[116],"SearchVL-RL-8k":[118],"RL.":[120],"Besides,":[121],"design":[123],"diverse":[125],"tool":[126,163],"environment":[127],"that":[128,160],"unifies":[129],"text":[130],"image":[132],"OCR,":[134],"cropping,":[135],"sharpening,":[136],"super-resolution,":[137],"perspective":[139],"correction,":[140],"combine":[144],"perception":[146],"external":[148],"knowledge":[149],"acquisition.":[150],"Finally,":[151],"propose":[153],"multi-turn":[155],"fatal-aware":[156],"GRPO":[157],"algorithm":[159],"handles":[161],"cascading":[162],"failures":[164],"by":[165],"masking":[166],"post-failure":[167],"tokens":[168],"while":[169],"preserving":[170],"useful":[171],"pre-failure":[172],"reasoning":[173],"one-sided":[175],"advantage":[176],"clamping.":[177],"Built":[178],"recipe,":[181],"OpenSearch-VL":[182],"delivers":[183],"substantial":[184],"performance":[185],"gains,":[186],"over":[188],"10-point":[189],"average":[190],"improvements":[191],"across":[192],"seven":[193],"benchmarks,":[194],"achieves":[196],"results":[197],"comparable":[198],"proprietary":[200],"commercial":[201],"several":[204],"tasks.":[205],"We":[206],"will":[207],"release":[208],"all":[209],"code,":[211],"support":[215],"research":[217],"agents.":[222]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-08T00:00:00"}