Medscape Reports Strong AI Benchmark Results, Details Proprietary ClinVal Framework

Performance on NOHARM v2 and HealthBench Professional benchmarks reflects Medscape's focus on clinical AI quality and safety, while its clinician-developed ClinVal framework supports ongoing evaluation and refinement.

NEWARK, N.J., Nov. 4, 2026 — Medscape put its underlying AI models to the test, scoring 88.6% on NOHARM v2 and 85.0% on HealthBench Professional, two leading benchmarks for clinical AI safety and performance. These strong results demonstrate Medscape's approach to clinical AI, combining model performance with the clinical context and trusted medical information healthcare professionals need.

Medscape is also sharing details of ClinVal, its proprietary evaluation framework, for the first time. ClinVal provides a rigorous methodology for assessing the quality of AI-generated answers and guiding further development of the Medscape AI answer engine.

Strong Results on Leading Clinical AI Benchmarks

Medscape's Clarus model scored 88.6% on NOHARM v2, a safety benchmark that examines the frequency and severity of potentially harmful errors in AI-generated medical recommendations. On HealthBench Professional, Medscape's Clarus model scored 85.0%. This benchmark assesses AI responses to a range of health-related scenarios against physician-defined standards for answer quality.

Compared with competing AI models with published results, Medscape's 88.6% score on NOHARM v2 exceeds the three highest reported competitor scores of 86.2%, 84.5% and 80.0%. On HealthBench Professional, Medscape's 85.0% score surpasses the three highest published competitor scores of 69.2%, 65.6% and 65.3%.

"These results reflect the rapid progress we're making in developing AI for healthcare," said Joshua Conrad, PharmD, Group VP Editorial & Content Strategy at Medscape. "We're laser-focused on systematic testing and continued refinement, with the goal of delivering meaningful value to healthcare professionals in real-world practice."

ClinVal: Advancing Continuous Clinical AI Evaluation and Improvement

Beyond benchmark testing, Medscape has developed ClinVal, its proprietary clinician-developed evaluation framework designed to systematically assess and improve the quality of answers generated by Medscape's AI models.

ClinVal evaluates AI-generated answers against an evidence-based set of 16 criteria across dimensions including safety, accuracy, relevance, and traceability; these were established by Medscape clinicians and refined in collaboration with AI engineering teams. The evaluation process combines automated AI assessments with clinician review, bringing clinical expertise into the assessment of answer quality.

Importantly, ClinVal supports a continuous cycle of evaluation and improvement. Findings help Medscape identify opportunities to enhance the product, and answers are re-evaluated following changes to assess their impact. This repeatable process enables Medscape to measure quality throughout the ongoing development of its AI technology, rather than relying solely on periodic external benchmark results.

Medscape has also shared ClinVal with external physicians to gather additional feedback on its evaluation methodology and clinical relevance.

"Strong benchmark results demonstrate what our AI can achieve today, but clinical AI quality requires an ongoing commitment to improvement," Conrad said. "With ClinVal, we've established a process that combines AI-driven evaluation with clinical expertise to identify where our answers can be better, make improvements, and measure the results. This continuous cycle is fundamental to how we're developing Medscape's AI models."

SOURCE Medscape