Wardstone
Wardstone is an LLM security API that sits between applications and language model providers, scanning inputs and outputs for threats across four categories in a single call: prompt attacks, content violations, data leakage, and unknown links.
It detects jailbreaks, prompt injections, harmful content (hate, violence, self-harm), PII (SSNs, credit cards, emails, phone numbers), and suspicious URLs. Each response returns risk bands per category with sub-30ms latency.
Works with any LLM provider. REST API with SDKs for TypeScript, Python, Go, Ruby, PHP, Java, and C#. Free tier at 10,000 calls/month, no credit card required. Includes a browser-based playground for testing.
Learn more
TROJAI
Even the best AI models can have hidden risks. Identify and address potential problems before they impact your business, ensuring smooth AI adoption and compliance. AI applications are vulnerable to new and sophisticated attacks. Stay ahead of the curve by protecting your models and applications from data poisoning, prompt injection, and other emerging threats. Leverage cutting-edge public AI services with confidence. We help you ensure responsible use and prevent data leaks, so you can focus on innovation without worry. The TROJAI security platform enables organizations to comply with benchmarks such as the OWASP AI framework as well as privacy regulations by testing models prior to deployment and protecting applications from things such as sensitive data loss once deployed.
Learn more
LangProtect
LangProtect is an AI-native security and governance platform that protects LLM and Generative AI applications from prompt injection, jailbreaks, sensitive data leakage, and unsafe or non-compliant outputs. Built for production GenAI, it enforces real-time runtime controls at the AI execution layer by inspecting prompts, model responses, and tool/function calls as they happen. This allows teams to block high-risk behavior before it reaches end users, triggers downstream actions, or exposes confidential data.
LangProtect integrates into existing LLM stacks via an API-first approach with minimal latency and supports cloud, hybrid, and on-prem deployments for enterprise security and data residency needs. It also secures modern architectures such as RAG pipelines and agentic workflows with policy-driven enforcement, continuous visibility, and audit-ready governance.
Learn more
WebOrion Protector Plus
WebOrion Protector Plus is a GPU-powered GenAI firewall engineered to provide mission-critical protection for generative AI applications. It offers real-time defenses against evolving threats such as prompt injection attacks, sensitive data leakage, and content hallucinations. Key features include prompt injection attack protection, safeguarding intellectual property and personally identifiable information (PII) from exposure, content moderation and validation to ensure accurate and on-topic LLM responses, and user input rate limiting to mitigate risks of security vulnerability exploitation and unbounded consumption. At the core of its capabilities is ShieldPrompt, a multi-layered defense system that utilizes context evaluation through LLM analysis of user prompts, canary checks by embedding fake prompts to detect potential data leaks, pand revention of jailbreaks using Byte Pair Encoding (BPE) tokenization with adaptive dropout.
Learn more
