Test AI security in your browser
The same protections that run on every Anoman API call — try them right now, free, without an account. English and Bahasa Indonesia.
The tools
Try a live guarded call
Run a REAL LLM call through the full guarded gateway. Watch prompt injection get blocked before the model, PII get masked, and see the cost in Rupiah on our Jakarta-hosted gateway — in one call.
Run a live call →
Guardrail tester
Paste any prompt (or a model's response) and see the guardrails run live — prompt injection, jailbreaks, PII, and unsafe content. The exact checks that guard every API call.
Test guardrails →
Faithfulness tester
Paste a source and an AI answer and see whether the answer is grounded in the source or hallucinated — a clear Grounded / Uncertain / Ungrounded verdict, calibrated for English and Bahasa.
Test faithfulness →
AI data-security diagnostic
Answer a few questions and get an AI-generated report of your top-5 AI data-security & UU PDP gaps, with practical fixes — sent to your inbox. For builders and founders.
Run the diagnostic →
Why they're free
See the product work before you sign up
Most gateways ask you to sign up, add billing, and write code before anything happens. These tools flip that: feel the guardrails, the groundedness check, and the residency receipt first — then get a free key and run it on your own traffic. Nothing you type in these tools is stored.
Ready to run it on your own traffic?
A free key includes up to 2M tokens/day on guarded models, free — plus a $3 credit and a premium trial. No card, billed in Rupiah, via our Jakarta gateway.