US NIST-built test platform for assessing trustworthy characteristics of AI models, providing a REST API, web interface and Python client to design, run and track reproducible experiments including adversarial red-team scenarios. Intended to support the Measure function of the AI Risk Management Framework rather than to certify systems.
open_source · generally available · Research snapshot 2026-09-06
Demonstrate reproducing an adversarial robustness experiment in Dioptra and exporting the tracked evidence for an internal AI risk review.
Capabilities and evidence
Support labels reflect the supplied research. Documentation and vendor claims are not independent product tests. “Not established” means the researcher did not find support; it does not prove a capability is absent.
Documented by provider
The repository states Dioptra is a software test platform for assessing trustworthy characteristics of AI that supports the Measure function of the NIST AI Risk Management Framework and provides a REST API, web interface and Python client for designing, executing and tracking experiments.
Limit: It measures characteristics for chosen experiments; it does not produce conformity statements or certifications.
NIST documentation describes Dioptra as a modular microservice environment for reproducible, trackable and reusable AI workflows to measure, analyse and track AI risks, with stated uses covering first-, second- and third-party model testing, research, evaluations and red-teaming.
Limit: Documentation does not state the license name on the page and does not quantify supported model scale.