Model Fingerprint

Verify whether a model matches its claimed identity

Configuration

Your API key is used only for this check and is not stored after the request completes.

Result

Formal Conclusion

Once a check starts, this panel shows the conclusion first, then the capability, prompt, and detailed diagnostic evidence.

Capability Probe

Capability comparison data is not available yet.

Prompt Probe

0 / 5

PromptStatusSimilarityScoreableErrorSummary
Prompt comparison data is not available yet.

Detailed Diagnostics

Detailed diagnostics will appear after evidence is available.

How it works

Combine standard fingerprint training with live response comparison to verify whether the target model matches its claimed identity.

Fingerprint training

Build standard fingerprints from large-scale official model responses, capturing stable signals across structure, capability, and style.

Fingerprint comparison

Send five high-information prompts to the target endpoint and compare the extracted signals against preset fingerprint models.

Conclusion output

Use completion coverage and match consistency to output similarity, confidence interval, and a final verdict. When evidence is limited, show only provisional observations.