{"article":{"annotation":{"article_id":"dualentry-accounting-ai-benchmark-results-2026","available_at":"2026-10-21T15:31:26.435811Z","geography_revision":"ef6ee0010dda5fe7b240fd514749b43e214c57da73222d2e40ec83b276dd0977","id":"ca665b60f4e564bec44bf25889d7aa413d030d98f41c3a0f8de54bf2c378c49f","input_sha256":"ef30cfe4eed102e1bb01a3c8974258649a7333a17763711175e4b86350468d15","limitations":["Rule-derived subjects are candidates, not verified editorial labels or calibrated model probabilities.","Coverage is distinct from event location. Unknown country/city/coordinates remain absent; no automatic geocoding.","Article geometries keep their stated precision and methodology; reproducing them is not independent coordinate verification."],"method":"controlled-rules-v1","places":[],"rules_sha256":"fb5ad3c27c91cb840f7a0a05f7f64a5cb26ad0374409ff9457f277c7092e22b4","schema_version":1,"subjects":[{"concept_id":"medtop:20001298","confidence":null,"confidence_method":null,"evidence":[{"field":"en.title+en.summary","language":"en","note":"Subject suggested from the title or summary: AI","source_urls":["https://markets.financialcontent.com/stocks/article/bizwire-2026-10-21-deepseeks-latest-update-scores-lower-on-accounting-work-as-open-models-trail-on-consistency-dualentry-benchmark-finds"]}],"method":"controlled-rules-v1","rule_id":"artificial-intelligence","status":"candidate"}],"taxonomy_revision":"1f2db31d811fca07e99b16d402c10e2cc2965c9df1c7b41d9be348b40a24e7fb","taxonomy_version":"IPTC Media Topics 2026-Q2"},"api_url":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026?lang=en","category":"technology","content":{"key_points":[],"paragraphs":["DualEntry announced results on 21 October from its Accounting AI Benchmark, which evaluates models on 101 synthetic cases across eight accounting areas. The company said each of 19 models was tested three times with the same accounting environment, tools and instructions; a task counted as passed when a model met the fixed scoring rule in at least two runs.","DualEntry reported Gemini 4 Argon at 91.1% overall accuracy and DeepSeek V4.1 Flash 1017 at 82.2%, compared with 84.2% for the earlier DeepSeek V4.1 Flash version. The company’s benchmark page says results use a versioned October 13 evaluation release. These scores describe this synthetic test set and methodology; they do not establish performance on live accounting work."],"sections":[]},"corrections":[],"country":null,"data":{"blocks":[{"description":{"ar":"قارنت الشركة 19 نموذجاً للذكاء الاصطناعي في 101 حالة محاسبية اصطناعية، وشغّلت كل حالة ثلاث مرات بالبيئة والتعليمات نفسها. وأفادت بأن Gemini 4 Argon سجل أعلى دقة في جولة الاختبار.","de":"Das Unternehmen verglich 19 KI-Modelle anhand von 101 synthetischen Buchhaltungsfällen, die jeweils dreimal in derselben Umgebung und mit denselben Anweisungen ausgeführt wurden. Gemini 4 Argon erzielte laut Bericht in dieser Runde die höchste Genauigkeit.","en":"The company compared 19 AI models on 101 synthetic accounting cases, each run three times under the same environment and instructions. It reported Gemini 4 Argon with the highest accuracy in that test round.","es":"La empresa comparó 19 modelos de IA en 101 casos contables sintéticos, ejecutados tres veces con el mismo entorno e instrucciones. Informó que Gemini 4 Argon obtuvo la mayor exactitud en esa ronda.","fr":"L’entreprise a comparé 19 modèles d’IA sur 101 cas comptables synthétiques, chacun exécuté trois fois dans le même environnement et avec les mêmes consignes. Elle a indiqué que Gemini 4 Argon affichait la meilleure exactitude lors de cette série.","pt":"A empresa comparou 19 modelos de IA em 101 casos contábeis sintéticos, cada um executado três vezes no mesmo ambiente e com as mesmas instruções. Segundo o relatório, o Gemini 4 Argon teve a maior precisão nessa rodada.","tr":"Şirket 19 yapay zekâ modelini, aynı ortam ve talimatlarla üç kez çalıştırılan 101 sentetik muhasebe vakasında karşılaştırdı. Bu test turunda en yüksek doğruluğu Gemini 4 Argon için bildirdi."},"id":"announcement","items":[{"date":"2026-10-21","description":{"ar":"قارنت الشركة 19 نموذجاً للذكاء الاصطناعي في 101 حالة محاسبية اصطناعية، وشغّلت كل حالة ثلاث مرات بالبيئة والتعليمات نفسها. وأفادت بأن Gemini 4 Argon سجل أعلى دقة في جولة الاختبار.","de":"Das Unternehmen verglich 19 KI-Modelle anhand von 101 synthetischen Buchhaltungsfällen, die jeweils dreimal in derselben Umgebung und mit denselben Anweisungen ausgeführt wurden. Gemini 4 Argon erzielte laut Bericht in dieser Runde die höchste Genauigkeit.","en":"The company compared 19 AI models on 101 synthetic accounting cases, each run three times under the same environment and instructions. It reported Gemini 4 Argon with the highest accuracy in that test round.","es":"La empresa comparó 19 modelos de IA en 101 casos contables sintéticos, ejecutados tres veces con el mismo entorno e instrucciones. Informó que Gemini 4 Argon obtuvo la mayor exactitud en esa ronda.","fr":"L’entreprise a comparé 19 modèles d’IA sur 101 cas comptables synthétiques, chacun exécuté trois fois dans le même environnement et avec les mêmes consignes. Elle a indiqué que Gemini 4 Argon affichait la meilleure exactitude lors de cette série.","pt":"A empresa comparou 19 modelos de IA em 101 casos contábeis sintéticos, cada um executado três vezes no mesmo ambiente e com as mesmas instruções. Segundo o relatório, o Gemini 4 Argon teve a maior precisão nessa rodada.","tr":"Şirket 19 yapay zekâ modelini, aynı ortam ve talimatlarla üç kez çalıştırılan 101 sentetik muhasebe vakasında karşılaştırdı. Bu test turunda en yüksek doğruluğu Gemini 4 Argon için bildirdi."},"label":{"ar":"DualEntry تنشر نتائج معيار اصطناعي للذكاء الاصطناعي في المحاسبة","de":"DualEntry veröffentlicht Ergebnisse eines synthetischen KI-Benchmarks für Buchhaltung","en":"DualEntry publishes results from synthetic accounting AI benchmark","es":"DualEntry publica resultados de un benchmark sintético de IA contable","fr":"DualEntry publie les résultats d’un benchmark synthétique d’IA comptable","pt":"A DualEntry publica resultados de benchmark sintético de IA contábil","tr":"DualEntry, sentetik muhasebe yapay zekâ kıyaslama sonuçlarını yayımladı"},"status":"occurred"}],"methodology":{"ar":"يعتمد على إعلان الشركة في 21 أكتوبر ولا يستنتج نتائج مستقلة.","de":"Grundlage ist die Mitteilung des Unternehmens vom 21. Oktober; unabhängige Ergebnisse werden nicht abgeleitet.","en":"Based on the company’s 21 October announcement; no independent outcome is inferred.","es":"Basado en el anuncio de la empresa del 21 de octubre; no se infieren resultados independientes.","fr":"Fondé sur l’annonce de l’entreprise du 21 octobre ; aucun résultat indépendant n’est déduit.","pt":"Baseado no anúncio da empresa de 21 de outubro; não são inferidos resultados independentes.","tr":"Şirketin 21 Ekim tarihli duyurusuna dayanmaktadır; bağımsız sonuç çıkarımı yapılmamıştır."},"sourceUrls":["https://markets.financialcontent.com/stocks/article/bizwire-2026-10-21-deepseeks-latest-update-scores-lower-on-accounting-work-as-open-models-trail-on-consistency-dualentry-benchmark-finds"],"title":{"ar":"DualEntry تنشر نتائج معيار اصطناعي للذكاء الاصطناعي في المحاسبة","de":"DualEntry veröffentlicht Ergebnisse eines synthetischen KI-Benchmarks für Buchhaltung","en":"DualEntry publishes results from synthetic accounting AI benchmark","es":"DualEntry publica resultados de un benchmark sintético de IA contable","fr":"DualEntry publie les résultats d’un benchmark synthétique d’IA comptable","pt":"A DualEntry publica resultados de benchmark sintético de IA contábil","tr":"DualEntry, sentetik muhasebe yapay zekâ kıyaslama sonuçlarını yayımladı"},"type":"events","updatedAt":"2026-10-21T15:09:28Z"}],"version":1},"data_downloads":[{"formats":{"csv":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026/data/announcement?lang=en&format=csv","json":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026/data/announcement?lang=en&format=json"},"id":"announcement"}],"data_types":["events"],"editorial_illustration":{"alt":"Conceptual illustration of technology, processors and connected information.","asset_id":"newsverum-v1-technology","avif_variants":[{"height":320,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-480.avif","width":480},{"height":480,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-720.avif","width":720},{"height":640,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-960.avif","width":960},{"height":800,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1200.avif","width":1200},{"height":1024,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.avif","width":1536}],"credit":"Newsverum","documentary":false,"height":1024,"kind":"symbolic-editorial-illustration","language":"en","scope":"topic","url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.webp","variants":[{"height":320,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-480.webp","width":480},{"height":640,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-960.webp","width":960},{"height":1024,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.webp","width":1536}],"width":1536},"event_id":"event-0749b010-2fe6-52fd-b6e7-a7325dd46028","event_period":"2026-10-21","event_period_basis":"event-or-reference-period","id":"dualentry-accounting-ai-benchmark-results-2026","image":null,"importance":0.6,"language":"en","modified_at":"2026-10-21T15:31:12+00:00","publication_basis":"publication-record","published_at":"2026-10-21T15:31:12+00:00","reading_minutes":1,"region":"world","section_illustration":{"alt":"Conceptual illustration of technology, processors and connected information.","asset_id":"newsverum-v1-technology","avif_variants":[{"height":320,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-480.avif","width":480},{"height":480,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-720.avif","width":720},{"height":640,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-960.avif","width":960},{"height":800,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1200.avif","width":1200},{"height":1024,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.avif","width":1536}],"credit":"Newsverum","documentary":false,"height":1024,"kind":"symbolic-editorial-illustration","language":"en","scope":"section","url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.webp","variants":[{"height":320,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-480.webp","width":480},{"height":640,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-960.webp","width":960},{"height":1024,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.webp","width":1536}],"width":1536},"slug":"dualentry-accounting-ai-benchmark-results-2026","sources":[{"checkedAt":"2026-10-21","name":"DualEntry","publishedAt":"2026-10-21","url":"https://markets.financialcontent.com/stocks/article/bizwire-2026-10-21-deepseeks-latest-update-scores-lower-on-accounting-work-as-open-models-trail-on-consistency-dualentry-benchmark-finds"}],"summary":"The company compared 19 AI models on 101 synthetic accounting cases, each run three times under the same environment and instructions. It reported Gemini 4 Argon with the highest accuracy in that test round.","text":"The company compared 19 AI models on 101 synthetic accounting cases, each run three times under the same environment and instructions. It reported Gemini 4 Argon with the highest accuracy in that test round.\n\nDualEntry announced results on 21 October from its Accounting AI Benchmark, which evaluates models on 101 synthetic cases across eight accounting areas. The company said each of 19 models was tested three times with the same accounting environment, tools and instructions; a task counted as passed when a model met the fixed scoring rule in at least two runs.\n\nDualEntry reported Gemini 4 Argon at 91.1% overall accuracy and DeepSeek V4.1 Flash 1017 at 82.2%, compared with 84.2% for the earlier DeepSeek V4.1 Flash version. The company’s benchmark page says results use a versioned October 13 evaluation release. These scores describe this synthetic test set and methodology; they do not establish performance on live accounting work.","title":"DualEntry publishes results from synthetic accounting AI benchmark","topics":[],"translations":[{"api_url":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026?lang=tr","language":"tr","url":"https://newsverum.com/tr/news/dualentry-accounting-ai-benchmark-results-2026"},{"api_url":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026?lang=en","language":"en","url":"https://newsverum.com/en/news/dualentry-accounting-ai-benchmark-results-2026"},{"api_url":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026?lang=ar","language":"ar","url":"https://newsverum.com/ar/news/dualentry-accounting-ai-benchmark-results-2026"},{"api_url":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026?lang=fr","language":"fr","url":"https://newsverum.com/fr/news/dualentry-accounting-ai-benchmark-results-2026"},{"api_url":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026?lang=es","language":"es","url":"https://newsverum.com/es/news/dualentry-accounting-ai-benchmark-results-2026"},{"api_url":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026?lang=de","language":"de","url":"https://newsverum.com/de/news/dualentry-accounting-ai-benchmark-results-2026"},{"api_url":"https://newsverum.com/api/v1/articles/dualentry-accounting-ai-benchmark-results-2026?lang=pt","language":"pt","url":"https://newsverum.com/pt/news/dualentry-accounting-ai-benchmark-results-2026"}],"url":"https://newsverum.com/en/news/dualentry-accounting-ai-benchmark-results-2026"},"version":"1"}