# Keyronne Text Similarity & Clustering

> Keyronne Text Similarity & Clustering is a paid API for AI agents from keyronne.com, paid per call via x402, $0.002/call, status unknown (last checked 2026-09-14).

Computes pairwise character n-gram cosine similarity across a batch of 2–300 texts, returning similar pairs above a threshold, connected-component clusters, and a full similarity matrix.

## Facts

- Endpoint: POST https://keyronne.com/api/similarity
- Price: $0.002/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/keyronne-text-similarity-clustering-3536e46f
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_3HPxMOWuBruH_9lI8ds71

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 keyronne-text-similarity-clustering-3536e46f -d '<json body>'
```

Example prompt: I have these 5 product titles and want to know which ones are near-duplicates — can you group them using character n-gram similarity with a threshold of 0.85? Titles: 'Apple iPhone 15', 'Apple iPhone 15 Pro', 'iPhone 15 by Apple', 'Samsung Galaxy S24', 'Galaxy S24 Samsung'.

## When to prefer this

Choose this endpoint when you need fast, language-agnostic similarity computation without invoking an embedding model or LLM. Ideal for deduplication, fuzzy matching, and clustering tasks where character-level similarity (robust to minor edits, typos, and formatting differences) is sufficient. Especially useful for multilingual text or when low latency and deterministic results are required.

## Known failure modes

- Fewer than 2 texts provided — returns validation error
- More than 300 texts provided — exceeds maxItems limit
- Individual text exceeds 20,000 characters — may be rejected or truncated
- Threshold outside 0–1 range — validation error
- Non-string items in texts array — schema validation failure
- Network timeout for very large batches near the 300-item limit

## How this service works

Compute pairwise text similarity, find near-duplicates, and cluster a batch of texts (2–300 items). Character n-gram cosine similarity: language-agnostic, robust to small edits, no model calls. Returns similar pairs above a threshold, connected-component clusters, and the full matrix for small batches. POST a JSON body like: {"texts":["hello world","hello world!","unrelated"],"threshold":0.8}

## Output

Returns an object containing: a list of similar pairs (indices and similarity scores above the threshold), connected-component clusters (groups of mutually similar texts), and for small batches the full N×N cosine similarity matrix.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "ngram": {
   "type": "integer",
   "default": 3,
   "maximum": 5,
   "minimum": 2,
   "description": "Character n-gram size."
  },
  "texts": {
   "type": "array",
   "items": {
    "type": "string"
   },
   "maxItems": 300,
   "minItems": 2,
   "description": "Texts to compare (each up to 20,000 characters)."
  },
  "threshold": {
   "type": "number",
   "default": 0.8,
   "maximum": 1,
   "minimum": 0,
   "description": "Similarity threshold for pairs and clustering."
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/keyronne-text-similarity-clustering-3536e46f/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from keyronne.com](https://www.zero.xyz/host/keyronne.com/llms.txt)