Tier System

Understanding ModelRed's subscription tiers and access levels for security testing capabilities

ModelRed uses a tier-based subscription model that provides access to increasingly sophisticated security testing capabilities. Each tier is designed to meet different security requirements and organizational needs.

Tier Overview

🆓

Free

Basic Security

14 test suites • 2 AI models • 10 assessments/month

Perfect for learning and small projects

POPULAR

🚀

Starter

$49/month

24 test suites • 10 AI models • 100 assessments/month

Ideal for growing teams and production apps

🔒

Pro

$149/month

13 test suites • 50 AI models • 500 assessments/month

Research-grade security for serious developers

🏢

Enterprise

Custom Pricing

7 test suites • Unlimited models • Unlimited assessments

Maximum security for large organizations

Free Tier

Perfect for individual developers and small projects

What's Included

14 Test Suites: Basic encoding attacks and security fundamentals
2 AI Models: Register up to 2 models for testing
10 Assessments/Month: Sufficient for development and testing
Community Support: Access to documentation and community forums

Focus Areas

Encoding & injection attacks (Base64, hex, ROT13, etc.)
Character encoding & obfuscation
Terminal & system attacks
Basic security validation

Use Cases

Learning AI security fundamentals
Personal projects and experimentation
Small-scale applications with basic security needs
Educational and research purposes

Starter Tier

Ideal for growing teams and production applications

What's Included

All Free Tier Features plus:
24 Additional Test Suites: Content safety, social engineering, advanced manipulation
10 AI Models: Scale up your testing infrastructure
100 Assessments/Month: Regular security validation
Email Support: Priority technical assistance

Advanced Capabilities

Content Safety Testing: Toxicity, hate speech, and harmful content detection
Social Engineering: Grandma attacks, emotional manipulation techniques
Prompt Injection: Advanced injection and jailbreaking attempts
Technical Vulnerabilities: Template injection, SQL injection testing
Reasoning & Misinformation: Hallucination and false assertion detection

Use Cases

Production applications with user-facing AI
Content moderation systems
Customer service chatbots
Educational AI platforms
Small to medium businesses

Pro Tier

For serious developers requiring research-grade security

What's Included

All Starter Tier Features plus:
13 Research-Grade Test Suites: State-of-the-art attack techniques
50 AI Models: Enterprise-scale model management
500 Assessments/Month: Continuous security validation
Priority Support: Dedicated technical support team

Research-Grade Features

Advanced Jailbreaking: DAN 11.0 ablation studies with 127 variations
AutoDAN Attacks: Gradient descent optimization jailbreaks
Latent Injection: Hidden prompt injection in documents and translations
Real-World Attacks: Curated attacks from actual adversarial interactions
Advanced Attack Techniques: GCG cached attacks, semantic exploration

Use Cases

High-stakes AI applications
Financial services and healthcare AI
Government and defense contractors
AI research and development teams
Security-critical applications

Enterprise Tier

For large organizations with the highest security requirements

What's Included

All Pro Tier Features plus:
7 Enterprise-Exclusive Test Suites: Malware and advanced threat detection
Unlimited AI Models: No restrictions on model registration
Unlimited Assessments: Continuous security monitoring
Dedicated Support: 24/7 enterprise support with SLA

Enterprise-Exclusive Features

Malware Generation Detection: Code generation vulnerabilities across languages
XSS & Data Exfiltration: Advanced web security and data protection
Supply Chain Attacks: Package hallucination across multiple ecosystems
Custom Integration: API customization and enterprise features

Enterprise Benefits

Custom Pricing: Volume discounts and flexible billing
On-Premise Deployment: Run ModelRed in your infrastructure
Training & Consulting: Security team training and threat modeling
Priority Feature Development: Influence product roadmap

SDK Usage Examples

Basic Usage

import asyncio
from modelred import ModelRed

async def main():
    async with ModelRed(api_key="mr_your_api_key_here") as client:
        # Free tier: Basic encoding tests
        result = await client.run_assessment(
            model_id="my-model",
            test_suites=["base64_injection", "hexadecimal_injection"],
            wait_for_completion=True
        )

        # Starter tier: Content safety
        result = await client.run_assessment(
            model_id="my-model",
            test_suites=["toxicity_generation", "lmrc_bullying"],
            priority="high",
            wait_for_completion=True
        )

asyncio.run(main())

Checking Tier Limits

# Check your current tier and usage
account_info = await client.get_account_info()
print(f"Current tier: {account_info.tier}")
print(f"Models registered: {account_info.models_used}/{account_info.models_limit}")
print(f"Assessments this month: {account_info.assessments_used}/{account_info.assessments_limit}")