Once a check starts, this panel shows the conclusion first, then the capability, prompt, and detailed diagnostic evidence.
Capability comparison data is not available yet.
0 / 5
| Prompt | Status | Similarity | Scoreable | Error | Summary |
|---|---|---|---|---|---|
| Prompt comparison data is not available yet. | |||||
Detailed diagnostics will appear after evidence is available.
Combine standard fingerprint training with live response comparison to verify whether the target model matches its claimed identity.
Build standard fingerprints from large-scale official model responses, capturing stable signals across structure, capability, and style.
Send five high-information prompts to the target endpoint and compare the extracted signals against preset fingerprint models.
Use completion coverage and match consistency to output similarity, confidence interval, and a final verdict. When evidence is limited, show only provisional observations.