Supported Models and Use Cases
A comprehensive evaluation framework that spans the entire AI development lifecycle

Machine Learning
- Data Drift
- Classification Rates
- Root Mean Square
- Precision & Recall
- Many More
Generative AI
- Hallucination Rates
- Data Security Controls
- Acceptable Use Policies
- Domain-specific Evals, inc. custom code
- Inference & hallucination count
- Pass & Fail rates for Toxicity, PII & Sensitive Data
- Tokens & Model cost





Agentic AI






Built for production AI
Whether you operate in the cloud, across multiple clouds, or inside tightly controlled infrastructure, Arthur deploys where your AI systems live.
From fast-start SaaS to fully air-gapped environments, we support the security, compliance, and performance requirements of modern AI teams.
On-Prem / Air-Gapped
Deploy Arthur entirely within your controlled infrastructure, including fully air-gapped environments with no external internet connectivity. Built for highly regulated industries such as financial services, defense, healthcare, and critical infrastructure.

AWS Deployment
Deploy Arthur directly into your AWS environment for full control over infrastructure, networking, and data residency. Arthur integrates natively with your existing AWS stack, including Bedrock, SageMaker, EKS, IAM, and VPC configurations.

GCP Deployment
Run Arthur inside your Google Cloud environment with seamless integration into your broader GCP ecosystem. Designed for teams building on Vertex AI, Kubernetes, or custom model pipelines inside GCP.
SaaS Deployment
Get started instantly with Arthur’s fully managed SaaS deployment.No infrastructure management required. Connect your models, agents, and data pipelines and begin evaluating, tracing, and governing production AI immediately.
Model Agnostic
Platform Agnostic
Flexible Deployment
See what Arthur can do for you.
