Simple, Transparent Pricing

From a free single-model test to full enterprise monitoring. Choose the plan that fits your AI safety needs.

25+ Models

GDPR Compliant

SOC2 Ready

14-Day Money-Back Guarantee

"UMMRO helped us identify critical safety gaps in our production AI models before deployment. The PDF report was exactly what our board needed."

— Security Lead, Fortune 500 Tech Company

Monthly Annual SAVE 20%

Quick Scan

Fast model assessment

$199 one-time

5 model audit, 24h delivery

5 models tested
Basic PDF report with risk grades
17 reframing strategies tested
Soft refusal detection overview
Branded PDF with logo
EU AI Act compliance mapping

Standard Audit

Complete AI safety audit

$499 one-time

Full 25+ model audit, one report

Full 25+ model audit
Branded PDF report (boardroom-ready)
EU AI Act compliance check
17 reframing strategies with effectiveness analysis
Comparative safety matrix (A-F grades)
Soft refusal detection and quantification
Prioritized remediation roadmap
Machine-readable summary.json
Email support (48h response)

Delivered within 48 hours

Deep Dive

Enterprise-grade security

$1,499 one-time

Comprehensive enterprise audit

Everything in Professional
API access (REST + JSON)
Continuous monitoring with drift alerts
Custom model configurations
Custom regulatory framework mapping
CI/CD pipeline integration
Priority support with SLA
Dedicated account manager
On-premise deployment option

Or contact sales for custom enterprise pricing

Compare Plans

Feature Comparison

Feature	Quick Scan	Standard	Deep Dive
Models tested	5	25+	25+ (extended analysis)
Reframing strategies	17	17	17 + Custom
Report format	Text	Branded PDF	Branded PDF
Risk grades (A-F)
EU AI Act compliance
Soft refusal detection	Overview	Full quantification	Full quantification
Remediation roadmap
API access
Continuous monitoring
Dedicated account manager
Support	Community	Email (48h)	Priority + SLA

FAQ

Frequently Asked Questions

What models do you test?

UMMRO tests 25+ production AI models across all major providers, including OpenAI (GPT-4o, GPT-4 Turbo, o1, o3), Anthropic (Claude Opus, Sonnet, Haiku), Google (Gemini 2.5 Pro, Flash), Meta (Llama 4), Mistral, DeepSeek, Cohere, and more. The Starter plan tests GPT-4o only. Professional and Enterprise plans cover the full model catalogue, and Enterprise clients can add custom or private models.

How long does an audit take?

The automated stress testing across all 25+ models completes in a matter of hours. The full certified PDF report, including human-reviewed analysis, risk grading, compliance mapping, and the remediation roadmap, is delivered within 48 hours of engagement start. The free Starter audit (1 model) typically completes within 24 hours.

What is in the PDF report?

The branded PDF report includes: an executive summary for leadership, risk grades A through F for each model tested, a comparative safety matrix, soft refusal detection rates and analysis, reframing effectiveness breakdowns for all 17 strategies, EU AI Act compliance mapping, regulatory cross-references (NIST AI RMF, ISO 42001, SOC 2), a prioritized remediation roadmap with actionable steps, and a machine-readable summary.json for CI/CD integration. Typically 30-50 pages depending on the number of models tested.

Is my data secure?

Absolutely. UMMRO operates as a black-box testing framework and never requires access to your model weights, training data, or proprietary source code. We only need API-level access or standard model identifiers. All audit data is encrypted in transit (TLS 1.3) and at rest (AES-256). We are GDPR-compliant, and audit results are retained only for the duration specified in your agreement. Enterprise clients can opt for on-premise deployment for full data sovereignty.

Can I customize which models to test?

Yes. The Professional plan audits the full 25+ model catalogue by default, but you can specify a subset if you only use certain models. Enterprise clients can add custom model endpoints, fine-tuned models, or private deployments to the audit harness. You define the scope, and we configure the testing framework to match your exact stack.

Do you offer volume discounts?

Yes. The subscription pricing (toggle above) provides a 20% discount compared to one-time purchases for recurring audits. For teams needing more than 5 audits per month or multi-department rollouts, our Enterprise plan offers custom volume pricing. Contact ahmed@ummro.dev to discuss your requirements.

What is your refund policy?

If you are not satisfied with your audit report, let us know within 14 days of delivery and we will either re-run the audit with adjusted parameters at no extra cost or issue a full refund. For subscription plans, you can cancel at any time and will retain access through the end of your current billing period. No questions asked.

How does continuous monitoring work?

Available on the Enterprise plan, continuous monitoring runs automated safety scans on a schedule you define (daily, weekly, or on model update events). When a model provider pushes an update that changes safety behavior, UMMRO detects the drift and sends you an alert with a delta report showing exactly what changed. This integrates with your CI/CD pipeline via our REST API and can trigger deployment gates automatically. You will never be caught off guard by a model safety regression.

Ready to audit your
AI safety posture?

Choose the right audit tier for your needs — from Quick Scan at $199 to comprehensive Deep Dive at $1,499.