# Model Substitution Attestation (Fingerprint & Verify)

> Model Substitution Attestation (Fingerprint & Verify) is a paid API for AI agents from pearl-valium-stickers-upc.trycloudflare.com, paid per call via x402, $0.1/call, status unknown (last checked 2026-10-31).

Detects silent model weight substitution at any OpenAI-compatible inference endpoint by fingerprinting a model's responses and later verifying whether the weights have changed.

## Facts

- Endpoint: POST https://pearl-valium-stickers-upc.trycloudflare.com/v1/attest/model?utm_source=zero.xyz
- Price: $0.1/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-31
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/model-substitution-attestation-fingerprint-verify-25d63886
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap__7z449eNwGSIEflXSXqSv

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability model-substitution-attestation-fingerprint-verify-25d63886 -d '<json body>'
```

Example prompt: Fingerprint the inference endpoint at https://api.example.com and record a proof of what model they're currently serving so I can verify later whether they swap it out.

## When to prefer this

Choose this endpoint when you need deterministic, evidence-based proof of model identity at an OpenAI-compatible inference endpoint — especially for auditing third-party LLM providers for silent model substitution. Unlike model-based classifiers or self-reported metadata, this endpoint uses behavioral fingerprinting with raw evidence, making it suitable for compliance, vendor accountability, or long-term monitoring workflows.

## Known failure modes

- Target endpoint unreachable or returns non-OpenAI-compatible responses
- Fingerprint record not found when attempting verify mode
- Endpoint rate-limits or blocks the probe queries
- Inconclusive verdict due to high variance in model responses
- Payment of 0.1 USDC not processed, blocking request
- Self-consistency score too low to produce reliable fingerprint

## How this service works

Proof-of-Model: detect silent model substitution at any OpenAI-compatible inference endpoint. mode=fingerprint records what a seller serves today; mode=verify re-probes later and proves whether the weights changed. Deterministic, evidence included, no model used to judge.

## Output

Returns an attestation record containing a kind field (fingerprint or verify), a claim about the model being served, a verdict on whether weights match, an array of probe comparisons, a mean similarity score, and a self-consistency score — all deterministically derived with raw evidence included, no LLM judge used.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "kind": {
   "type": "string"
  },
  "claim": {
   "type": "string"
  },
  "verdict": {
   "type": "string"
  },
  "comparisons": {
   "type": "array"
  },
  "mean_similarity": {
   "type": "number"
  },
  "self_consistency_now": {
   "type": "number"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/model-substitution-attestation-fingerprint-verify-25d63886/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from pearl-valium-stickers-upc.trycloudflare.com](https://www.zero.xyz/host/pearl-valium-stickers-upc.trycloudflare.com/llms.txt)