AI Hallucination Risk Checker
A free, browser-based tool. Runs entirely in your browser — no sign up, nothing stored.
Your prompt
Rule-based, runs in your browser — nothing is uploaded.
How to use this calculator
- Paste the prompt you plan to send to an AI.
- Read the Low / Medium / High risk rating.
- Apply the suggested fix for each flag before you send it.
How it works
Certain requests make models more likely to fabricate: anything needing live/current data, exact figures or statistics, verbatim quotes or citations, predictions about the future, or high-stakes legal/medical/financial topics. The checker flags these patterns and weights them into a risk level.
It can't tell whether any specific answer is true — it spots the question patterns that tend to produce confident but unverifiable answers, and suggests how to reduce the risk (provide the facts, ask for sources, say 'if unsure, say so').
Worked example
“current population of Tokyo and the latest stock price today”. This trips both the current-information and exact-figure flags, landing in the Medium–High band — the fix is to supply the data or require sources.
Common mistakes
- Asking for an exact statistic or quote without giving the source — the model may invent a plausible-looking one.
- Treating a confident tone as proof of accuracy. Verify anything important.
Frequently asked questions
Does this use AI or send my prompt anywhere?
No. It runs entirely in your browser using rule-based checks — your prompt is never uploaded or sent to any AI model. Still, avoid pasting passwords, API keys or private data into any tool.
Does a Low rating mean the answer will be correct?
No — it means your prompt avoids the common triggers for invented answers. Always sanity-check facts regardless of the rating.
Related tools
Tip: Enter any known values to calculate the remaining results.
All calculations run in your browser. Your inputs are never saved or transmitted.