{"schema_version":1,"kind":"jevbench-api-full-addenda","entries":[{"key":"microsoft-decision-1","published_at":null,"prepared_at":"2026-10-10T05:10:18.790981+00:00","publication_status":"approved_for_publication","release":"CR-401 Microsoft-Decision-1 full native API addendum","provenance":{"parent":"v1.6-fastlane-20261009","method":"O1S","g_med_s":7.088097744249485,"counts":{"P":300,"S":1200},"cost_n":1479,"input_sha256":"6833e921ae525a328a4cf199a3e8145fa657fe2c9a9fe520e2953aab6da60ee6","run_sha256":"cadf99e94de455277604759b31028cfb14663f3b815555e5ea7909e0bb891c27","scorer_sha256":"83398383da9d011cc30236d0a241a9c125fc5efbbfc041f5564e7c5f699d896e","categories_sha256":"47a4db8923dc9edadcd74ee357c852099a7473e8f6b99100e8fe61767b9e63c5","acceptance_sha256":"1a21b7d132e77af0196edebb1c4f9060e2eb3067aa04ec5ed99cb94bdbc808ce","cost_mask_sha256":"50d582d3976ba79d458756f931764cf6e80bc623fbbe2fa611d8744752e3fc7b","source_url":"https://benchmarkheaven.com/api/jevbench/api-addenda","release_authorization_sha256":"ea7e7d50d31ba80c47bdba27f9b32f35f03dabfd2bb1fe6264f6e9d5588023a3"},"row":{"api_flag":true,"author":"Microsoft","axes":{"calibration":84.39873071167106,"cost":61.452075716621465,"intelligence":57.16958510591244,"speed":81.54517543213291},"calibration":{"parts":{"choice":{"ece_hard":0.09636174358194516,"ece_score":80.72765128361097,"mean_tvd":null,"n":735,"n_ece_hard":307,"n_tvd":0,"score":80.72765128361097},"noul":{"abstention_rate":0.2278820375335121,"brier":0.10202697897728127,"ece":0.05824671413738715,"n":373,"score":88.35065717252257},"score":{"interval90_coverage":0.9056603773584906,"interval90_width":0.21495956873315356,"n":371,"nrps":0.07658334569212163,"score":84.11788367887968,"top_ece":0.12052949036514232}},"score":84.39873071167106},"capability":70.78415790879176,"capability_ci95":[65.79924499904618,73.0808555865328],"composite_ci95":{"A":[66.02436050159598,70.40957825472715],"B":[62.32949160874876,68.09460025172818],"C":[46.743685735358696,70.27466026550705]},"cost":{"basis":"Advertised Azure Foundry tariff .042/M input, output free; genuine native usage, billed invoice unknown","common_basis":"{\"n_items\": 1479, \"rule\": \"common\", \"usd_per_1000_all_items\": 0.019002368000000002}","kind":"tariff","usage_estimated_rows":0,"usd_per_1000":0.0192721784989858},"display":"Microsoft-Decision-1 (Azure Foundry)","endpoint_kind":"api","intelligence":{"I_open":65.54484720922025,"I_sealed":49.68435126148425,"base":57.61459923535225,"excess":8.772398203486514,"gap":15.860495947735998,"penalty":0.9922760179651349,"per_type_split":{"open|choice":{"cc":82.42557001849185,"n":{"easy":33,"hard":68,"judge":32,"standard":17},"tiers":{"easy":100.00000000000001,"hard":75.46398238439761,"judge":84.17930125247199,"standard":84.93093344495604}},"open|noul":{"cc":51.46376811594203,"n":{"easy":14,"hard":23,"judge":20,"standard":18},"tiers":{"easy":100,"hard":47.82608695652173,"judge":30.000000000000004,"standard":66.66666666666667}},"open|score":{"cc":62.74520349322689,"n":{"easy":9,"hard":29,"judge":21,"standard":16},"tiers":{"easy":51.19290622964672,"hard":67.62102226028924,"judge":74.3930948447687,"standard":41.29787756357959}},"sealed|choice":{"cc":65.77933402862911,"n":{"easy":73,"hard":254,"judge":141,"standard":132},"tiers":{"easy":85.18366087045993,"hard":59.4822734946539,"judge":65.37585421412302,"standard":69.27651139742319}},"sealed|noul":{"cc":37.62824744344246,"n":{"easy":48,"hard":116,"judge":67,"standard":69},"tiers":{"easy":54.16666666666667,"hard":25.86206896551724,"judge":31.343283582089555,"standard":62.318840579710155}},"sealed|score":{"cc":45.6454723123812,"n":{"easy":36,"hard":111,"judge":74,"standard":79},"tiers":{"easy":68.67980276922125,"hard":32.88780129611539,"judge":46.813302889415894,"standard":57.89190325094079}}}},"jevbench_score":69.11289175277149,"key":"microsoft-decision-1","last_measured_on":"2026-10-10","listing":"ranked","model_pin":{"deployment":"decision-1","model":"Microsoft-Decision-1","provider":"Microsoft/Azure Foundry","response_model":"microsoft-decision-1","version":"1"},"ranked":true,"repo":"https://commandline.microsoft.com/microsoft-decision-1-model-foundry/","scores":{"A":69.11289175277149,"B":66.34102888286164,"C":62.74608663211473},"speed":{"adjustment":"none (API)","n":124,"p50_s_adjusted":0.45925578847527504,"p50_s_raw":0.45925578847527504,"p95_s_adjusted":1.5255548182874916,"p95_s_raw":1.5255548182874916},"status":{"answered_ok":1479,"missing":0,"rows":1500,"status":"complete"},"v16":{"api_subset":false,"api_subset_tag":null,"breakdowns_equated":false,"ci95":{"B":1000,"C":[79.38267975134342,86.54667450947431],"I":[50.44641980227999,60.43490499481045],"capability":[65.79924499904618,73.0808555865328],"composite":{"A":[66.02436050159598,70.40957825472715],"B":[62.32949160874876,68.09460025172818],"C":[46.743685735358696,70.27466026550705]},"seed":16},"ci95_n":1200,"complete":true,"equated":false,"equating_flag":null,"equating_offset":null,"full_set_api":true,"g_med_used":7.088097744249485,"lane":"api","metric_basis":"S u P (full set, API lane, v1.6.1 amendment; not equated)","n_items":1500,"n_public":300,"n_sealed_side":1200,"on_S":{"calibration":84.39873071167106,"capability":70.78415790879176,"intelligence":57.16958510591244},"per_type":{"choice":{"accuracy":0.7653333333333333,"calibration":80.72765128361097,"cc":69.61967049617368,"cc_by_type":{"choice":69.61967049617368},"cc_tiered":74.10245202356049,"cc_tiered_open":82.42557001849185,"cc_tiered_sealed":65.77933402862911,"low_n":false,"n":750,"n_by_type":{"choice":750},"score":69.61967049617368},"noul":{"accuracy":0.7173333333333334,"calibration":88.35065717252257,"cc":43.466666666666676,"cc_by_type":{"noul":43.466666666666676},"cc_tiered":44.546007779692246,"cc_tiered_open":51.46376811594203,"cc_tiered_sealed":37.62824744344246,"low_n":false,"n":375,"n_by_type":{"noul":375},"score":43.466666666666676},"score":{"accuracy":0.6933333333333334,"calibration":84.11788367887968,"cc":51.165347833026,"cc_by_type":{"score":51.165347833026},"cc_tiered":54.19533790280404,"cc_tiered_open":62.74520349322689,"cc_tiered_sealed":45.6454723123812,"low_n":false,"n":375,"n_by_type":{"score":375},"score":51.165347833026}},"raw_on_A":null,"run_sha256":"cadf99e94de455277604759b31028cfb14663f3b815555e5ea7909e0bb891c27","set":"S+P"},"support":{"choice":"native","noul":"native","score":"native"},"noul_decisive":{"supported":true,"decisive_rate":0.7721179624664879,"acc_among_decisive":0.9340277777777778,"valid":373}},"coverage":{"languages":[{"answered_ok":1260,"competence":56.83,"coverage_n":1281,"errors":21,"key":"en","label":"English","n":1281,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":97,"competence":51.36,"coverage_n":97,"errors":0,"key":"de","label":"German","n":97,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":80,"competence":49.75,"coverage_n":80,"errors":0,"key":"fr","label":"French","n":80,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":94,"competence":46.21,"coverage_n":94,"errors":0,"key":"es","label":"Spanish","n":94,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":79,"competence":45.37,"coverage_n":79,"errors":0,"key":"it","label":"Italian","n":79,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":85,"competence":58.75,"coverage_n":85,"errors":0,"key":"pt","label":"Portuguese","n":85,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":78,"competence":46.12,"coverage_n":78,"errors":0,"key":"nl","label":"Dutch","n":78,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":77,"competence":49.39,"coverage_n":77,"errors":0,"key":"da","label":"Danish","n":77,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":73,"competence":42.01,"coverage_n":73,"errors":0,"key":"sv","label":"Swedish","n":73,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":64,"competence":32.59,"coverage_n":64,"errors":0,"key":"no","label":"Norwegian","n":64,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":72,"competence":23.96,"coverage_n":72,"errors":0,"key":"fi","label":"Finnish","n":72,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":96,"competence":36.65,"coverage_n":96,"errors":0,"key":"pl","label":"Polish","n":96,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":73,"competence":38.58,"coverage_n":73,"errors":0,"key":"cs","label":"Czech","n":73,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":73,"competence":33.36,"coverage_n":73,"errors":0,"key":"el","label":"Greek","n":73,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":74,"competence":42.68,"coverage_n":74,"errors":0,"key":"tr","label":"Turkish","n":74,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":70,"competence":38.45,"coverage_n":70,"errors":0,"key":"uk","label":"Ukrainian","n":70,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":76,"competence":44.21,"coverage_n":76,"errors":0,"key":"ar","label":"Arabic","n":76,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":85,"competence":33.8,"coverage_n":85,"errors":0,"key":"hi","label":"Hindi","n":85,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":68,"competence":53.79,"coverage_n":68,"errors":1,"key":"zh","label":"Chinese","n":69,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":85,"competence":54.97,"coverage_n":85,"errors":0,"key":"ja","label":"Japanese","n":85,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":75,"competence":33.1,"coverage_n":75,"errors":0,"key":"ko","label":"Korean","n":75,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":70,"competence":57.3,"coverage_n":70,"errors":0,"key":"id","label":"Indonesian","n":70,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":79,"competence":88.32,"coverage_n":79,"errors":0,"key":"mixed","label":"Mixed-language","n":79,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]}],"topics":[{"answered_ok":448,"competence":20.68,"coverage_n":459,"errors":11,"key":"math","label":"Math & numbers","n":459,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":277,"competence":57.43,"coverage_n":277,"errors":0,"key":"coding","label":"Coding & software","n":277,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":1599,"competence":53.18,"coverage_n":1599,"errors":1,"key":"law_policy","label":"Rules, policy & law","n":1600,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":236,"competence":56.25,"coverage_n":236,"errors":0,"key":"finance_commerce","label":"Finance & commerce","n":236,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":157,"competence":56.25,"coverage_n":167,"errors":10,"key":"support_ops","label":"Support & operations","n":167,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":152,"competence":75.13,"coverage_n":152,"errors":0,"key":"everyday_language","label":"Everyday language","n":152,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":114,"competence":53.57,"coverage_n":114,"errors":0,"key":"safety_security","label":"Safety & security","n":114,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]}],"usecases":[{"answered_ok":97,"competence":85.32,"coverage_n":97,"errors":0,"key":"search_retrieval","label":"Search and retrieval","n":97,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":97,"competence":36.91,"coverage_n":97,"errors":0,"key":"scientific_discovery","label":"Scientific discovery","n":97,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":400,"competence":64.3,"coverage_n":400,"errors":0,"key":"model_routing","label":"Model routing","n":400,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":101,"competence":69.68,"coverage_n":101,"errors":0,"key":"llm_guardrails","label":"LLM guardrails","n":101,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":96,"competence":19.83,"coverage_n":96,"errors":0,"key":"code_linting","label":"Semantic code linting","n":96,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":96,"competence":36.89,"coverage_n":96,"errors":0,"key":"feature_extraction","label":"Feature extraction for predictive modeling","n":96,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":98,"competence":29.7,"coverage_n":98,"errors":0,"key":"recruiting","label":"Recruiting","n":98,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":114,"competence":63.85,"coverage_n":114,"errors":0,"key":"lead_generation","label":"Lead generation","n":114,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":240,"competence":60.37,"coverage_n":249,"errors":9,"key":"customer_support","label":"Customer support","n":249,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":107,"competence":42.83,"coverage_n":107,"errors":0,"key":"insurance_claims","label":"Insurance claims","n":107,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":115,"competence":28.96,"coverage_n":115,"errors":1,"key":"financial_crime","label":"Financial crime","n":116,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":471,"competence":63.42,"coverage_n":471,"errors":0,"key":"legal_compliance","label":"Legal and compliance","n":471,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":114,"competence":52.63,"coverage_n":114,"errors":0,"key":"ecommerce_marketplaces","label":"E-commerce marketplaces","n":114,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":98,"competence":67.15,"coverage_n":98,"errors":0,"key":"moderation_trust_safety","label":"Moderation and trust and safety","n":98,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":107,"competence":37,"coverage_n":107,"errors":0,"key":"advertising","label":"Advertising","n":107,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":90,"competence":25.06,"coverage_n":90,"errors":0,"key":"gaming","label":"Gaming","n":90,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":126,"competence":51.21,"coverage_n":126,"errors":0,"key":"risk_assessment","label":"Risk assessment","n":126,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":88,"competence":0,"coverage_n":88,"errors":0,"key":"demand_forecasting","label":"Demand forecasting","n":88,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":98,"competence":35.49,"coverage_n":98,"errors":0,"key":"knowledge_graphs","label":"Graphs and knowledge graphs","n":98,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]},{"answered_ok":230,"competence":47.1,"coverage_n":242,"errors":12,"key":"other","label":"Other","n":242,"pool":"S1200+P300+L1+L2+L3; full native Foundry union","types":["choice","noul","score"]}]}}]}