The faster, simpler alternative
for LLM security
Wardstone Guard detects prompt injections, jailbreaks, harmful content, and PII leakage in a single API call. Sub-30ms. No complex setup.
Works with any LLM
How Wardstone compares
Everything you need in one tool, not five.
| Feature | Wardstone | Others |
|---|---|---|
| Prompt injection detection | Yes | Varies |
| Jailbreak detection | Yes | Some |
| PII / data leakage detection | Yes | Often separate tool |
| Content moderation | Yes | Often separate tool |
| Unknown link detection | Yes | No |
| Latency | Sub-30ms | 100-500ms |
| Free tier | 10,000 calls/mo | Limited or none |
| SDKs | 7 languages | 1-3 languages |
| Setup complexity | Single API call | Multi-step config |
All-in-one detection
Four threat categories in a single API call. No stitching together multiple tools.
Built for speed
ONNX-optimized inference on CPU. Sub-30ms P95 latency, fast enough for synchronous pipelines.
Drop-in integration
REST API with SDKs for TypeScript, Python, Go, Ruby, PHP, Java, and C#. Ship in minutes.
Integrate in minutes
A few lines of code is all it takes.
TypeScript
import Wardstone from 'wardstone';
const wardstone = new Wardstone({ apiKey: 'YOUR_API_KEY' });
const result = await wardstone.detect({ text: userInput });
if (result.flagged) {
// Block the request
}Switch to Wardstone in minutes
Free up to 10,000 API calls/month. No credit card required.