# Foundry OCR

> Foundry OCR is a paid API for AI agents from minty-sprout.onrender.com, paid per call via x402, $0.006/call, status unknown (last checked 2026-11-01).

Extracts all text from an image or PDF URL and returns it as Markdown using vision-based OCR

## Facts

- Endpoint: POST https://minty-sprout.onrender.com/v1/ocr?utm_source=zero.xyz
- Price: $0.006/call
- Payment: x402
- Status: unknown
- Last checked: 2026-11-01
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/foundry-ocr-9abf1dee
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_bf331hx65jd9VlN6uSWZg

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 foundry-ocr-9abf1dee -d '<json body>'
```

Example prompt: Can you extract all the text from this receipt image for me? Here's the URL: https://example.com/receipt.jpg

## When to prefer this

Choose this endpoint when you need to extract text from a publicly accessible image URL (PNG, JPEG, WebP) or a text-based PDF URL. It is ideal for receipts, invoices, screenshots, scanned documents, and any image containing readable text. Prefer this over general-purpose vision APIs when you need clean Markdown output at a low fixed cost per call ($0.006 USDC) and your source is already hosted at a URL rather than a local file.

## Known failure modes

- Invalid or inaccessible URL returns an error — ensure the URL is publicly reachable
- Unsupported file format (e.g. non-text PDF with complex layout) may yield incomplete or empty text
- Very large images or PDFs may time out or return partial results
- Handwritten or low-resolution images may result in poor OCR accuracy
- Non-image, non-PDF URLs will likely return an error or empty text

## How this service works

OCR: extract all text from an image URL (photos, screenshots, scans, receipts, invoices) as Markdown; text PDFs read directly. Input: an image or PDF URL.

## Output

A JSON object containing the original URL, the full extracted text as a Markdown string, the OCR method used (e.g. 'vision-ocr'), and the character count of the extracted text.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "input": {
   "type": "string",
   "maxLength": 8000,
   "description": "URL of a PNG, JPEG, WebP or text PDF"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "url": "https://www.base.org/images/base-open-graph-v2.png",
  "text": "base\n\nThe blockchain for global finance.",
  "method": "vision-ocr",
  "characters": 38
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/foundry-ocr-9abf1dee/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from minty-sprout.onrender.com](https://www.zero.xyz/host/minty-sprout.onrender.com/llms.txt)