Cached prefix replays return identical text with TTFT close to 0 and inter-token times near zero variance. A proxy that aggressively caches is not strictly substituting models, but it is also not running fresh inference.
Algorithm
Send the same prompt twice with temperature > 0. If the responses are identical and TTFT drops by ≥ 50% on the second call, mark as cached.
Thresholds
Condition
Verdict contribution
Identical text + TTFT drop ≥ 50%
Cache hit
Identical text + TTFT drop 20–50%
Possible cache
Otherwise
No cache observed
Limitations
Identical text at temperature > 0 across two calls is not impossible — short, deterministic prompts can produce stable outputs.
Anysingle signal cannot provemalicious behavior. Proxies may show anomalies for legitimate reasons (regional routing, A/B testing, degradation strategies, cache optimization).
Token ratio deviation may result from ChatML wrapping, system prompt injection, or tokenizer version differences — not necessarily intentional inflation.
Model identity judgment is based on statistical fingerprint matching, not cryptographic proof. Quantization, fine-tuning, and post-processing can all alter fingerprints.
MMD distribution tests are sensitive to temperature, sampling parameters, and system prompts. Significant p-values mean distributional difference, not proof of substitution.
Logprobs unavailability is increasingly common (many providers disable it by default in 2025-2026) and does not by itself indicate deception.
ITT rhythm fingerprinting is an early-stage technique. Network jitter, TCP coalescing, and gateway buffering can produce false signals.
This tool generates reference-grade evidence chains, not legal conclusions. Do not make definitive accusations based solely on this report.
The wording in the report refers to statistical "deviations" or "signal inconsistencies". Please do not use this to make fraud or deception claims against any service provider.