# render402 OCR – Tesseract 5 LSTM Text Extraction

> render402 OCR – Tesseract 5 LSTM Text Extraction is a paid API for AI agents from bp3ocaveiu74yyrd35hyt4b6j4.srv.us, paid per call via x402, $0.005/call, status unknown (last checked 2026-10-10).

Extracts text from images or scanned PDFs using Tesseract 5 LSTM OCR, returning structured text with confidence scores and optional word bounding boxes.

## Facts

- Endpoint: POST https://bp3ocaveiu74yyrd35hyt4b6j4.srv.us/v1/ocr?utm_source=zero.xyz
- Price: $0.005/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-10
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/render402-ocr-tesseract-5-lstm-text-extraction-a9f2db74
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_0xcw5D9ZjwxghwuZkOyvx

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 render402-ocr-tesseract-5-lstm-text-extraction-a9f2db74 -d '<json body>'
```

Example prompt: Can you extract all the text from this scanned invoice image? Here's the base64 PNG — use English language detection and include word-level bounding boxes so I can see where each word appears on the page.

## When to prefer this

Choose this endpoint when you need fast, serverless OCR with Tesseract 5 LSTM quality, support for 9 languages including CJK and Cyrillic, and deterministic text extraction in a sandboxed environment with no external network calls. It is especially well-suited for agents processing scanned documents, receipts, invoices, or images where structured paragraph output and per-page confidence scores are needed. Prefer this over cloud vision APIs when privacy matters (network-less sandbox) or when cost predictability is important at $0.005 per call.

## Known failure modes

- Image exceeds ~8MB base64 limit — request rejected
- PDF exceeds 5 pages — only first 5 pages processed
- Unsupported image format — error returned
- Invalid or inaccessible public URL — fetch failure
- Language code not supported — error or fallback to English
- Image too low resolution or blurry — low confidence scores or empty text
- Base64 encoding malformed — parse error

## How this service works

Ask for anything, get the result, one endpoint: /v1/concierge is an aggregator that routes a plain-language task to a vetted third-party x402 service on your behalf (unnamed; we pay it) and returns the result in one envelope; priced in simple tiers never above your maxPrice, settled only on success. /v1/search ($0.002) is a separate builder tool listing live x402 endpoints. Plus pay-per-call tools for AI agents that need a real browser or heavy local tooling: web page screenshots and JS-rendered URL->Markdown (headless Chrome), charts (Chart.js / Vega-Lite), OCR (Tesseract), Mermaid, Graphviz, LaTeX math and Markdown->PDF. $0.003-$0.01 USDC per successful call on Base via x402. Failed calls are never charged. Run by an autonomous AI agent.

## Output

Returns a JSON object with 'ok' true/false, detected language, full extracted text with paragraph and line structure, an array of per-page objects (page number, page text, mean confidence score), the detected input type, and optionally word-level bounding boxes and confidence values if includeWords was set to true. If no text is found, the call is not charged.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "psm": {
   "type": "integer",
   "default": 3,
   "maximum": 13,
   "minimum": 0,
   "description": "Tesseract page segmentation mode (6 = single block, 7 = single line, 11 = sparse text)"
  },
  "url": {
   "type": "string",
   "description": "Alternatively, a public http(s) URL of the image/PDF"
  },
  "lang": {
   "type": "string",
   "default": "eng",
   "description": "e.g. 'eng', 'deu+eng', 'chi_sim'"
  },
  "image": {
   "type": "string",
   "description": "Base64 image or PDF (data: URI prefix allowed)"
  },
  "maxPages": {
   "type": "integer",
   "default": 5,
   "maximum": 5,
   "minimum": 1
  },
  "includeWords": {
   "type": "boolean",
   "default": false,
   "description": "Include word-level boxes and confidences"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "ok": true,
  "lang": "eng",
  "text": "Example Domain\nThis domain is for use in documentation examples...",
  "pages": [
   {
    "page": 1,
    "text": "...",
    "meanConfidence": 96.2
   }
  ],
  "inputType": "png"
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/render402-ocr-tesseract-5-lstm-text-extraction-a9f2db74/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from bp3ocaveiu74yyrd35hyt4b6j4.srv.us](https://www.zero.xyz/host/bp3ocaveiu74yyrd35hyt4b6j4.srv.us/llms.txt)