# Pocket Network Named Entity Recognition

> Pocket Network Named Entity Recognition is a paid API for AI agents from agent.pocket.network, paid per call via x402, $0.005/call, status unknown (last checked 2026-11-16).

Extracts named entities (persons, organisations, places, dates, money, percentages, emails, URLs, products) from English text using rule-based pattern matching, returning character offsets for each match.

## Facts

- Endpoint: POST https://agent.pocket.network/v1/named-entity-recognition?utm_source=zero.xyz
- Price: $0.005/call
- Payment: x402
- Status: unknown
- Last checked: 2026-11-16
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/pocket-network-named-entity-recognition-47f82ac5
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_1x52Q6igvq2V-Cbi-ywpT

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 pocket-network-named-entity-recognition-47f82ac5 -d '<json body>'
```

Example prompt: Can you run named entity recognition on this paragraph and tell me all the people, organisations, places, dates, and money amounts mentioned, along with where each one appears in the text: 'Apple CEO Tim Cook met with President Biden in Washington D.C. on June 5th to discuss a $3.5 billion investment deal.'

## When to prefer this

Choose this endpoint when you need fast, deterministic, rule-based NER on English text without setting up an account or API key, and you are comfortable paying $0.005 USDC per call. It is ideal for structured entity types (dates, money, emails, URLs, percentages) where pattern matching excels. Prefer a trained ML-based NER service if your text is informal, uncapitalised, multilingual, or requires contextual disambiguation of ambiguous entity names.

## Known failure modes

- Text too long — input exceeds 100,000 character limit, causing rejection
- Low recall on uncapitalised or informal text — rule-based system misses entities not matching capitalisation patterns or absent from word lists
- False positives on ambiguous tokens that match patterns but are not true entities
- No model-based disambiguation — cannot resolve whether 'Apple' is a company or the fruit based on context
- Non-English text returns poor or no results — system is tuned for English only
- Payment failure — insufficient USDC balance causes 402 response before processing

## How this service works

Rule-based named-entity recognition for English text, from patterns and word lists rather than a trained model: persons, organisations, places, dates, money, percentages, emails, URLs and products, with character offsets. Best on clean, well-capitalised text; it misses names off its lists. POST /v1/ner with {text}. Pay per request in USDC; no account, no API key.

## Output

A list of entity annotations, each containing the entity type (person, organisation, place, date, money, percentage, email, URL, product), the matched text span, and the start and end character offsets within the input text. Rule-based matching means results are deterministic but may miss names not on the built-in word lists; best accuracy on clean, well-capitalised text.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "text": {
   "type": "string",
   "maxLength": 100000,
   "description": "English text to tag."
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/pocket-network-named-entity-recognition-47f82ac5/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agent.pocket.network](https://www.zero.xyz/host/agent.pocket.network/llms.txt)
