LLM Benchmarking for HTR on French Historical Death Records

LLM Benchmarking for HTR on French Historical Death Records | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article LLM Benchmarking for HTR on French Historical Death Records Gabriel Frossard, Flavian Theurel, Wissam Al-Kendi, Franck Gechter This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9353651/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 6 You are reading this latest preprint version Abstract This study evaluates the performance of recent multimodal large language models (LLMs) for the transcription of historical handwritten and printed documents, with a focus on sources commonly used in historical demography. We benchmarked gpt-4o , several Qwen models, and other widely used LLMs against traditional OCR/HTR systems, including Tesseract and Ocelus, on a dataset of archival records containing complex layouts, marginal annotations, and variable handwriting styles. gpt-4o achieved the highest transcription accuracy across both paragraph text and marginal entries, while Qwen emerged as the most effective fully free and unrestricted alternative. Our analysis also highlights the limits of common evaluation metrics: the BLEU score did not consistently correlate with the Character Error Rate (CER), due to BLEU’s sensitivity to n-gram overlap rather than character-level fidelity. This distinction is critical for historical sources where spelling preservation and textual exactness are essential. Overall, the results demonstrate the growing potential of LLM-based approaches for large-scale historical data extraction and emphasize the need for continued benchmarking and domain-specific evaluation frameworks as new models rapidly emerge. Large Language Model Handwritten Text Recognition Historical Demography Archival Document Analysis Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 30 Apr, 2026 Reviewers agreed at journal 21 Apr, 2026 Reviewers invited by journal 21 Apr, 2026 Editor assigned by journal 20 Apr, 2026 Submission checks completed at journal 09 Apr, 2026 First submitted to journal 08 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9353651","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":627220859,"identity":"cece9e48-51be-4955-aa0c-2c1784769f61","order_by":0,"name":"Gabriel Frossard","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBklEQVRIiWNgGAWjYFACHgaGBAYLBvYGCJexn4GBDUjLEdIiwcBzAKplZgMzSIsxfi0MyFo2HCCgRd797LEPDyqAWqQPH/74o+ae7OYb+cce3WAwyMelxfBMXvKMhDNALXxpadI8x4qNt91IZjfOYTCwbMClpSHHmCGxTYLBnofHjJmBLSERqIVNOofhjwFOW/rfALX8A9rCw2P88ce/hMTNM8BaDHBqkZcA2dIA1mIgwduWkLhBgoAWA4l3yQwJxySAOtjSpHn7EoxnnHlsbpxjgMeW/tzDjD9qbOR4eJiBIfYtQba/PfHZ45wKPLYcgNA86OK4NABtacAtNwpGwSgYBaMAAgAMcEy/g3RIxwAAAABJRU5ErkJggg==","orcid":"","institution":"University of Technology of Belfort-Montbéliard","correspondingAuthor":true,"prefix":"","firstName":"Gabriel","middleName":"","lastName":"Frossard","suffix":""},{"id":627220860,"identity":"57c4da11-844a-4174-a498-31f3ffe9d1ba","order_by":1,"name":"Flavian Theurel","email":"","orcid":"","institution":"University of Technology of Belfort-Montbéliard","correspondingAuthor":false,"prefix":"","firstName":"Flavian","middleName":"","lastName":"Theurel","suffix":""},{"id":627220861,"identity":"d025939e-eede-4a80-b529-b45d2be75341","order_by":2,"name":"Wissam Al-Kendi","email":"","orcid":"","institution":"University of Technology of Belfort-Montbéliard","correspondingAuthor":false,"prefix":"","firstName":"Wissam","middleName":"","lastName":"Al-Kendi","suffix":""},{"id":627220862,"identity":"46e4cf40-6f8b-42b6-a0a4-dceea511bc01","order_by":3,"name":"Franck Gechter","email":"","orcid":"","institution":"University of Technology of Belfort-Montbéliard","correspondingAuthor":false,"prefix":"","firstName":"Franck","middleName":"","lastName":"Gechter","suffix":""}],"badges":[],"createdAt":"2026-04-08 08:08:55","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9353651/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9353651/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108491438,"identity":"3b34c2f3-86e0-4571-80e5-abe0f373de18","added_by":"auto","created_at":"2026-05-05 09:53:55","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5505898,"visible":true,"origin":"","legend":"","description":"","filename":"GFROSSARDmanuscriptJDH.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9353651/v1_covered_76090fe4-beed-48ba-a492-65802e03196f.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"LLM Benchmarking for HTR on French Historical Death Records","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-digital-humanities","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"ijdh","sideBox":"Learn more about [International Journal of Digital Humanities](http://link.springer.com/journal/42803)","snPcode":"42803","submissionUrl":"https://submission.nature.com/new-submission/42803/3","title":"International Journal of Digital Humanities","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Large Language Model, Handwritten Text Recognition, Historical Demography, Archival Document Analysis","lastPublishedDoi":"10.21203/rs.3.rs-9353651/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9353651/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study evaluates the performance of recent multimodal large language models (LLMs) for the transcription of historical handwritten and printed documents, with a focus on sources commonly used in historical demography. We benchmarked \u003cspan type=\"SmallCaps\" class=\"SmallCaps\" name=\"Emphasis\"\u003egpt-4o\u003c/span\u003e, several \u003cspan type=\"SmallCaps\" class=\"SmallCaps\" name=\"Emphasis\"\u003eQwen\u003c/span\u003e models, and other widely used LLMs against traditional OCR/HTR systems, including Tesseract and Ocelus, on a dataset of archival records containing complex layouts, marginal annotations, and variable handwriting styles. \u003cspan type=\"SmallCaps\" class=\"SmallCaps\" name=\"Emphasis\"\u003egpt-4o\u003c/span\u003e achieved the highest transcription accuracy across both paragraph text and marginal entries, while \u003cspan type=\"SmallCaps\" class=\"SmallCaps\" name=\"Emphasis\"\u003eQwen\u003c/span\u003e emerged as the most effective fully free and unrestricted alternative. Our analysis also highlights the limits of common evaluation metrics: the BLEU score did not consistently correlate with the Character Error Rate (CER), due to BLEU\u0026rsquo;s sensitivity to n-gram overlap rather than character-level fidelity. This distinction is critical for historical sources where spelling preservation and textual exactness are essential. Overall, the results demonstrate the growing potential of LLM-based approaches for large-scale historical data extraction and emphasize the need for continued benchmarking and domain-specific evaluation frameworks as new models rapidly emerge.\u003c/p\u003e","manuscriptTitle":"LLM Benchmarking for HTR on French Historical Death Records","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-30 13:57:27","doi":"10.21203/rs.3.rs-9353651/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-04-30T18:27:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"255652057999781966272033108360382254440","date":"2026-04-21T17:05:48+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-21T17:03:01+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-20T12:17:23+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-09T05:28:31+00:00","index":"","fulltext":""},{"type":"submitted","content":"International Journal of Digital Humanities","date":"2026-04-08T07:53:23+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-digital-humanities","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"ijdh","sideBox":"Learn more about [International Journal of Digital Humanities](http://link.springer.com/journal/42803)","snPcode":"42803","submissionUrl":"https://submission.nature.com/new-submission/42803/3","title":"International Journal of Digital Humanities","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"3683c068-fda6-49c3-8e26-02c8a33a00fd","owner":[],"postedDate":"April 30th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-04-30T18:27:59+00:00","index":6,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-30T13:57:27+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-30 13:57:27","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9353651","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9353651","identity":"rs-9353651","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}