Boards by clinical job

Physician-selected workplace tasks, from care consults to documentation and research, graded on physician-written rubrics.

OpenAI, 525 tasks

Expert-level multiple-choice questions over clinical images across 17 specialties.

Tsinghua University, 2,000 multimodal questions

A composite of clinical benchmarks spanning diagnostic and management reasoning, safety, multimodal imaging, and agentic capability.

ARISE AI Research Network, 6-benchmark composite

Holistic clinical evaluation across 121 tasks in a clinician-validated taxonomy, ranked by mean win rate.

Stanford CRFM, 121 tasks

Quality and compliance of SOAP notes generated from clinical visits, scored against documentation rubrics.

Vals AI, 100 SOAP-note cases

ICD-10-CM coding of whole hospital stays from discharge summaries and notes, checked against professional coders.

Vals AI, 2,755 patient records

Agents carrying out long-horizon physician workflows inside real EHR systems, verified by execution against those systems.

academic team, 100 clinical tasks

Agentic clinical reasoning over MIMIC-IV records through SQL and Python, at patient and population level.

academic team, 3,915-task test set

Agent harnesses completing realistic terminal-based healthcare tasks built from real clinical artifacts.

Microsoft Research, 54 agentic tasks

Long-horizon healthcare operations workflows for agents: prior authorization, utilization management, and care management.

actAVA, 75 operations workflows

Computer-use agents completing healthcare administration workflows: prior authorizations, denial appeals, and DME ordering.

Kinetic Systems, 135 admin tasks

Tests whether models apply and withdraw medication-safety rules correctly as patient information changes.

Hou et al., %

How often, and how severely, model consultation recommendations contain potentially harmful errors.

Stanford/Harvard consortium, 1,100 consultation cases