Honest fetching, three depths.
Turn a public product page into clean structured data - photos, title, description, variants, price - as JSON, CSV, or straight into your systems via API.
Read this first. ForgeFetch reads pages a visitor can already see. It does not bypass paywalls, logins, or bot-walls, and it respects robots.txt and per-domain rate limits. Supported sites grow continually; unknown sites return best-effort and say so. Overpromised "undetectable" tools burn their users - honesty is the point.
Light
Free forever
- ✓ Photos + title + photo links
- ✓ Well-formed, polite fetch of public pages
- ✓ JSON & CSV download
- ✓ Fair-use daily quota
Max full data
$8/mo
- ✓ Everything in Light
- ✓ Description, price, variants, specs
- ✓ Honest headless render (reads the page's own data)
- ✓ Variant formula for your listing sheet
- ✓ Higher daily quota + API key
Max Pro + your proxy
$25/mo
- ✓ Everything in Max
- ✓ Plug in your own proxy for reliability
- ✓ Largest daily quota
- ✓ Priority queue
After payment you land on a page that shows your API key. Paste it into the tool's key field, pick Max or Max Pro, and fetch. Honest note: Max Pro's proxy is your own - we never run proxy farms to defeat blocks.
The API
One endpoint. Send a URL and the fields you want; get clean JSON back. Keys are per user and gate quota by tier.
curl -X POST https://api.forgefetch.com/api/fetch \
-H "Content-Type: application/json" \
-H "X-API-Key: YOUR_KEY" \
-d '{
"url": "https://pl.aliexpress.com/item/100500.html",
"fields": ["photos","title","variants","price"],
"mode": "light"
}'
Response (trimmed):
{
"ok": true,
"confidence": "high",
"robots": { "allowed": true, "note": "robots.txt allows this path." },
"data": {
"title": "Folding Pocket Knife Outdoor",
"price": "PLN 12.34",
"photos": ["https://.../a.jpg", "https://.../b.jpg"],
"variant_formula": "size: 4-5 cm = 37 | 5-6,5 cm = 55 | 6,5-7 cm = 80",
"variant_confidence": "high"
}
}
CSV instead of JSON: POST the same body to /api/fetch.csv.
Honest limits
- ●Supported sites today, more continually. AliExpress works now. Others return best-effort with a confidence flag - never "every site instantly."
- ●We do not defeat access controls. No paywall, login, or bot-wall bypass. If a page needs payment or sign-in, ForgeFetch stops there.
- ●We respect robots.txt. Disallowed paths are not fetched. We throttle per domain so no site gets hammered.
- ●Low-confidence variants are flagged. Structures differ by site and template; a human reviews low-confidence reads before listing.
- ●Rendering renders, it doesn't disguise. Pro's browser mode reads JS-built public pages; it does not hide that it is automated or route around blocks.