AI Safety & Agents Tier 2 Ready POST /v1/hallucination-diff

LLM Grounding & Hallucination Check API — Fast First Filter for Numbers & Entities

Fast first filter to catch unsupported numbers, dates, and names in LLM answers vs RAG sources. Fast first-pass RAG grounding check that catches unsupported numbers, dates, percentages, and proper nouns in LLM answers against source text.

Try It Live in Your Browser · No API Key Required

1. Interactive Sandbox & Sample JSON Output

200 OK · Ready

Editable Request JSON (`POST /v1/hallucination-diff`)

Live 200 OK Response Envelope

{
  "ok": true,
  "data": {
    "grounded_score": 4,
    "verdict": "partially_grounded",
    "summary": {
      "total_claims": 1,
      "supported_claims": 0,
      "partial_claims": 1,
      "unsupported_claims": 0,
      "missing_numbers": [
        "24%"
      ],
      "missing_entities": [
        "Berlin"
      ]
    },
    "claims": [
      {
        "index": 0,
        "text": "Acme Corp generated $4.2M in Q3 2026 (24% YoY growth) and launched in Berlin.",
        "status": "partial",
        "overlap_score": 0.442,
        "missing_numbers": [
          "24%"
        ],
        "missing_entities": [
          "Berlin"
        ],
        "evidence": {
          "source_index": 0,
          "quote": "In Q3 2026, Acme Corp reported revenue of $4.2M and 18% year-over-year growth."
        }
      }
    ],
    "judge": "disabled",
    "ai_judge": null
  },
  "meta": {
    "request_id": "req_hd_01",
    "ms": 1
  }
}

Get API Key on RapidAPI Get API Key on Zyla Hub Download OpenAPI 3.0.3 Spec

2. Quickstart Code Examples

Full TypeScript & Python SDK →

curl -X POST "https://api.tinplate.dev/v1/hallucination-diff" \
  -H "Content-Type: application/json" \
  -H "X-RapidAPI-Key: YOUR_RAPIDAPI_KEY" \
  -d '{"sources":["In Q3 2026, Acme Corp reported revenue of $4.2M and 18% year-over-year growth."],"answer":"Acme Corp generated $4.2M in Q3 2026 (24% YoY growth) and launched in Berlin."}'

import { TinplateClient } from "@tinplate/sdk"; // npm install @tinplate/sdk

const tinplate = new TinplateClient({
  rapidApiKey: process.env.RAPIDAPI_KEY!,
});

const { data, meta } = await tinplate.call("/v1/hallucination-diff", {
  "sources": [
    "In Q3 2026, Acme Corp reported revenue of $4.2M and 18% year-over-year growth."
  ],
  "answer": "Acme Corp generated $4.2M in Q3 2026 (24% YoY growth) and launched in Berlin."
});
console.log(`[${meta.ms} ms]`, data);

import requests

response = requests.post(
    "https://api.tinplate.dev/v1/hallucination-diff",
    headers={
        "Content-Type": "application/json",
        "X-RapidAPI-Key": "YOUR_RAPIDAPI_KEY",
    },
    json={
  "sources": [
    "In Q3 2026, Acme Corp reported revenue of $4.2M and 18% year-over-year growth."
  ],
  "answer": "Acme Corp generated $4.2M in Q3 2026 (24% YoY growth) and launched in Berlin."
}
)
print(response.json())

const response = await fetch("https://api.tinplate.dev/v1/hallucination-diff", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "X-RapidAPI-Key": "YOUR_RAPIDAPI_KEY",
  },
  body: JSON.stringify({
  "sources": [
    "In Q3 2026, Acme Corp reported revenue of $4.2M and 18% year-over-year growth."
  ],
  "answer": "Acme Corp generated $4.2M in Q3 2026 (24% YoY growth) and launched in Berlin."
}),
});
const result = await response.json();
console.log(result);

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "tinplate_hallucination_diff",
    "arguments": {
      "sources": [
        "In Q3 2026, Acme Corp reported revenue of $4.2M and 18% year-over-year growth."
      ],
      "answer": "Acme Corp generated $4.2M in Q3 2026 (24% YoY growth) and launched in Berlin."
    }
  }
}

OpenAPI 3.0.3 Schema Contract

Request Parameters & Response Schema Reference

Request Body Parameters (`POST /v1/hallucination-diff`)

Parameter	Type & Constraints	Description
`answer`required	`string` minLen: 1, maxLen: 50000	Generated LLM answer to verify against source documents.
`sources`required	`string[]`	Grounding source passages or RAG chunks.
`strictness`optional	`"lenient" \| "balanced" \| "strict"`	Threshold profile for classifying claims as supported, partial, or unsupported.
`use_ai_judge`optional	`boolean`	Enable optional Cloudflare Workers AI second-pass semantic grounding judge (default false). When true, up to 4,000 chars of sources and 1,500 chars of answer are sent over Cloudflare's internal binding to @cf/meta/llama-3.1-8b-instruct (adds ~150-400 ms; zero model training or retention).

Response `data` Object Fields (`200 OK`)

Field	Type	Description
`grounded_score`	`integer`	Field 'grounded_score'
`verdict`	`"grounded" \| "partially_grounded" \| "ungrounded"`	Field 'verdict'
`summary`	`object`	Field 'summary'
`claims`	`object[]`	Field 'claims'
`judge`	`"disabled" \| "workers_ai" \| "unavailable"`	Field 'judge'
`ai_judge`	`object \| null`	Field 'ai_judge'

3. What Deterministic Hallucination Diff API Detects & Validates

Cheap, fast first filter (< 2 ms deterministic) that cross-checks numbers, percentages, dates, and proper-noun entities against RAG sources
Normalizes comma-formatted numbers (e.g. 10,000 vs 10000) to avoid false alarms on equivalent figures
Per-sentence unigram + bigram overlap scoring with best-matching source evidence quotes
Computes a 0–100 grounded_score and three-state verdict (grounded, partially_grounded, ungrounded)
Optional Workers AI semantic second pass (use_ai_judge: true) when semantic contradiction detection is needed

Honest Engineering Boundaries

4. Known Limitations & Test Verification

Release Version: v1.0.7 · Last Verified / Benchmarked: 14 November 2026 · Latency Note: Sub-millisecond figures measure in-Worker V8 compute time (typical client HTTPS network RTT to the nearest Cloudflare PoP is 15–65 ms) · See Full Portfolio Benchmarks & Corpora →

Benchmark v1.0.4 (14 November 2026, 25 labeled RAG source-answer pairs): 92.0% agreement with human judgment (23/25), 100.0% recall on fabricated numbers & proper-noun entities (13/13), 100.0% precision on grounded answers (10/10), P50 0.051 ms / P99 1.297 ms in deterministic mode.
Deterministic mode is a token and entity overlap filter, not a truth detector: it catches unsupported numbers, dates, and names, and cannot judge abstract logical reasoning, heavy paraphrase, or pure antonym/negation swaps ('never' -> 'always') unless use_ai_judge: true is enabled.
Typically under 5 ms in deterministic mode (use_ai_judge: false); enabling use_ai_judge: true adds ~150–400 ms for Workers AI inference.

5. Marketplace Plans & Payload Quotas

Plan	Monthly Price	Included Requests	Rate Limit	Max Request Payload	Best For
BASIC	$0.00 / month	500 / month (hard limit)	10 req / min	64 KB	Evaluation, local development & CI smoke tests
PRO	$9.99 / month	15,000 / month	60 req / min	128 KB	Solo developers & production apps
ULTRA	$29.99 / month	100,000 / month	300 req / min	256 KB	Growing SaaS & multi-agent pipelines
MEGA	$79.99 / month	500,000 / month	1,000 req / min	256 KB	High-volume enterprise workloads

Subscribe on RapidAPI Subscribe on Zyla Hub Full Pricing Details

6. Frequently Asked Questions

What types of hallucinations does deterministic mode catch vs miss?

Deterministic mode catches fabricated numbers, altered percentages/prices, wrong dates, and invented company/city/person names (100% recall on benchmark). Because it uses token and entity overlap, it cannot judge abstract reasoning, heavy paraphrase, or pure negation swaps ('is permitted' -> 'is not permitted')—it is a cheap first filter, not a truth detector.

What happens to my text when use_ai_judge: true is enabled?

By default use_ai_judge is false and text never leaves the Worker isolate. When use_ai_judge: true is set, an excerpt (up to 4,000 chars of sources and 1,500 chars of answer) is passed to Cloudflare Workers AI (@cf/meta/llama-3.1-8b-instruct) for a stateless semantic contradiction check with zero payload logging or model training.

How fast is it?

Typically under 5 ms in deterministic mode (P50 0.051 ms, P99 1.30 ms on benchmark). When use_ai_judge: true is enabled, Workers AI inference adds ~150–400 ms.

Zero Payload Logging

7. Data Handling, Privacy & Marketplace Proxy Disclosure

In default deterministic mode (use_ai_judge: false), all token/entity diffing runs in volatile V8 memory with zero external calls or payload logging. When use_ai_judge: true is explicitly passed, up to 4,000 chars of sources and 1,500 chars of answer are sent over Cloudflare's internal binding to Cloudflare Workers AI (@cf/meta/llama-3.1-8b-instruct) for stateless inference; neither Tinplate nor Cloudflare trains models on or retains Workers AI inputs/outputs.

Worker Zero-Logging Guarantee: Cloudflare Workers Observability and invocation logging are disabled in configuration (observability.enabled = false, head_sampling_rate = 0, invocation_logs = false). Payloads are processed strictly in volatile V8 memory and never logged or stored by Tinplate.

Marketplace Proxies vs. Direct Bearer Key: Requests sent via RapidAPI or Zyla API Hub pass through those marketplaces' proxy infrastructure (subject to their respective privacy policies). Security-sensitive workloads requiring end-to-end zero-intermediary processing can connect directly to https://api.tinplate.dev using a Direct Bearer Key (Authorization: Bearer <API_KEY>). Read our full Privacy Notice and OpenAPI 3.0.3 JSON Schema.