{"snapshot":{"schemaVersion":"1.0.0","snapshotId":"medi-bench-v1-2026-05-27","benchmarkVersion":"V1","title":"Medi Bench V1","publicationStatus":"archived","displayLabel":"ARCHIVED / DIRECTIONAL","capturedOn":"2026-05-27","publishedAt":null,"summary":"An archived display of the four rounded May 2026 comparisons for grounded-answer rate, cost per accepted answer, and p95 response speed.","provenance":"Values are transcribed from Compliance Health Technologies’ May 2026 Medi Workflow Bench chart.","metricPublication":{"status":"archived_display_only","comparativeMetricsPublished":true,"disclosure":"Only the rounded score, cost, and p95 values displayed on the archived chart are included."},"pilotProcess":null,"history":{"supersedes":null},"dataset":{"promptCount":null,"aggregateOnly":true,"disclosure":"The archived chart did not publish corpus size, repeats, confidence intervals, secondary rates, or prompt-level material."},"trackDefinitions":{"oracleContext":{"label":"Oracle-context","description":"V1 did not publish a separate oracle-context track."},"endToEnd":{"label":"End-to-end","description":"Archive slot for the single Medi workflow comparison displayed in V1; V1 did not publish separate track definitions."}},"chart":{"defaultTrack":"endToEnd","yAxis":"Grounded answer rate","xAxis":"Cost per accepted answer (USD)","bubble":"Larger bubbles indicate faster p95 latency","legacyMetricNote":"V1 reported grounded-answer rate, cost per accepted answer, and p95 latency. It is archived directional evidence with no published corpus size, repeats, or confidence intervals, and is not directly comparable with the planned V3 weighted methodology."},"modelEligibility":[{"modelId":"gpt-5.2","label":"gpt-5.2","tier":"archived","status":"completed","reason":"Displayed on the archived May 2026 chart."},{"modelId":"gpt-5.4-mini","label":"gpt-5.4-mini","tier":"archived","status":"completed","reason":"Displayed on the archived May 2026 chart."},{"modelId":"gpt-5.4","label":"gpt-5.4","tier":"archived","status":"completed","reason":"Displayed on the archived May 2026 chart."},{"modelId":"gpt-5-mini","label":"gpt-5-mini","tier":"archived","status":"completed","reason":"Displayed on the archived May 2026 chart."}],"models":[{"modelId":"gpt-5.2","label":"gpt-5.2","tier":"archived","eligibility":"eligible","eligibilityReason":"Displayed on the archived May 2026 chart.","tracks":{"oracleContext":{"status":"not_run","metrics":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidenceIntervals":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidence":{"level":"not_calculated","coverage":null,"method":"Oracle-context was not run in V1."},"run":{"completedAt":null,"promptCount":null,"repeatCount":null,"pricingDate":null,"configuration":"Not configured for the archived V1 fixture."}},"endToEnd":{"status":"completed","metrics":{"qualityScore":0.99,"costPerAcceptedUsd":0.004,"p95LatencyMs":9700},"confidenceIntervals":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidence":{"level":"directional","coverage":null,"method":"The archived aggregate fixture did not publish confidence intervals."},"run":{"completedAt":null,"promptCount":null,"repeatCount":null,"pricingDate":null,"configuration":"The archived chart did not publish run configuration."}}}},{"modelId":"gpt-5.4-mini","label":"gpt-5.4-mini","tier":"archived","eligibility":"eligible","eligibilityReason":"Displayed on the archived May 2026 chart.","tracks":{"oracleContext":{"status":"not_run","metrics":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidenceIntervals":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidence":{"level":"not_calculated","coverage":null,"method":"Oracle-context was not run in V1."},"run":{"completedAt":null,"promptCount":null,"repeatCount":null,"pricingDate":null,"configuration":"Not configured for the archived V1 fixture."}},"endToEnd":{"status":"completed","metrics":{"qualityScore":1,"costPerAcceptedUsd":0.0008,"p95LatencyMs":8100},"confidenceIntervals":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidence":{"level":"directional","coverage":null,"method":"The archived aggregate fixture did not publish confidence intervals."},"run":{"completedAt":null,"promptCount":null,"repeatCount":null,"pricingDate":null,"configuration":"The archived chart did not publish run configuration."}}}},{"modelId":"gpt-5.4","label":"gpt-5.4","tier":"archived","eligibility":"eligible","eligibilityReason":"Displayed on the archived May 2026 chart.","tracks":{"oracleContext":{"status":"not_run","metrics":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidenceIntervals":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidence":{"level":"not_calculated","coverage":null,"method":"Oracle-context was not run in V1."},"run":{"completedAt":null,"promptCount":null,"repeatCount":null,"pricingDate":null,"configuration":"Not configured for the archived V1 fixture."}},"endToEnd":{"status":"completed","metrics":{"qualityScore":0.98,"costPerAcceptedUsd":0.0035,"p95LatencyMs":10600},"confidenceIntervals":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidence":{"level":"directional","coverage":null,"method":"The archived aggregate fixture did not publish confidence intervals."},"run":{"completedAt":null,"promptCount":null,"repeatCount":null,"pricingDate":null,"configuration":"The archived chart did not publish run configuration."}}}},{"modelId":"gpt-5-mini","label":"gpt-5-mini","tier":"archived","eligibility":"eligible","eligibilityReason":"Displayed on the archived May 2026 chart.","tracks":{"oracleContext":{"status":"not_run","metrics":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidenceIntervals":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidence":{"level":"not_calculated","coverage":null,"method":"Oracle-context was not run in V1."},"run":{"completedAt":null,"promptCount":null,"repeatCount":null,"pricingDate":null,"configuration":"Not configured for the archived V1 fixture."}},"endToEnd":{"status":"completed","metrics":{"qualityScore":0.95,"costPerAcceptedUsd":0.0005,"p95LatencyMs":14300},"confidenceIntervals":{"qualityScore":null,"costPerAcceptedUsd":null,"p95LatencyMs":null},"confidence":{"level":"directional","coverage":null,"method":"The archived aggregate fixture did not publish confidence intervals."},"run":{"completedAt":null,"promptCount":null,"repeatCount":null,"pricingDate":null,"configuration":"The archived chart did not publish run configuration."}}}}],"methodology":{"modelEligibility":["V1 includes only the four models displayed on the archived May 2026 chart.","No later or untested model is added retroactively to the V1 record."],"oracleContextTrack":["Oracle-context was not part of the archived V1 fixture and all V1 entries remain not_run for this track."],"endToEndTrack":["V1 published one aggregate workflow comparison and did not disclose separate oracle-context and end-to-end tracks.","The four archived values are stored in the end-to-end display slot only to render the historical chart."],"scoring":["V1 reported grounded-answer rate, cost per accepted answer, and p95 latency.","The archive uses only the exact rounded values visible in the May 2026 chart; missing information remains null."],"confidenceIntervals":["V1 did not publish confidence intervals or repeat counts, so those fields remain null.","V1 is directional comparison evidence, not a safety certification."],"privacy":["Only aggregate values are included. Prompt text, responses, source content, and identifying context are forbidden from public snapshots."]},"limitations":["The archived chart does not publish corpus size, repeat count, confidence intervals, eligibility evidence, or secondary quality rates.","The displayed values are rounded and should not be interpreted as hidden-precision measurements.","Model, retrieval, pricing, and infrastructure conditions can change; this V1 fixture is intentionally not live.","The benchmark supports engineering comparison, not medical advice, legal advice, or final compliance decisions."],"changeLog":[{"date":"2026-05-27","version":"V1","summary":"Archived the four rounded model comparisons displayed in the May 2026 Medi Workflow Bench chart."}]},"integrity":{"algorithm":"Ed25519","keyId":"medi-bench-public-2026-07","contentSha256":"5e934b4a901d511467af10907a2f102e1046a57e8d75614e63ebbd82e7802078","signature":"xS48GZkBtVMVcZsRWbwq45/ZdbBtWrWxUekdNLrGXLjgiuNB5iTx+NcjA5GfGsUmucOx7K4QaECclOM9y11wDA==","verified":true}}