Skip to main content
Migrate in minutes

Already on Firecrawl? Swap the base URL, keep your code.

SuperScraper speaks the same core endpoints and shapes, so most code ports directly. Point your client at us and you are running. Below is the endpoint mapping and a straight, ground-truthed comparison. Where we are thinner, we say so.

The whole migration

- const client = new Firecrawl({ apiKey: process.env.FIRECRAWL_KEY });
+ const client = new Firecrawl({
+   apiKey: process.env.SUPERSCRAPER_KEY,
+   apiUrl: "https://superscraper-production-1381.up.railway.app",
+ });

// the rest of your code is unchanged
const { data } = await client.scrape("https://example.com");

Same call sites, same response shapes. Just a new base URL and key.

Endpoint mapping

FirecrawlSuperScraper
POST /v1/scrapePOST /v1/scrape(same formats & actions)
POST /v1/crawlPOST /v1/crawl(include/exclude paths + webhook)
GET /v1/crawl/:idGET /v1/jobs/:id(job status)
POST /v1/mapPOST /v1/map
POST /v1/extractPOST /v1/extract(schema-based)
POST /v1/batch/scrapePOST /v1/batch
screenshot (format)POST /v1/screenshot(also a dedicated endpoint)
POST /v1/searchPOST /v1/search

Core endpoints

CapabilitySuperScraperFirecrawl
Scrape (markdown / HTML / links / JSON / screenshot)
Crawl with include/exclude paths + webhook
Map / URL discovery
Batch (sync ≤100 + async)
Schema-based extract (single + multi-URL)
Dedicated screenshot endpoint(Firecrawl ships it as a format)
Browser actions (click/scroll/write/press/wait)
Search + scrape

Output & extraction

CapabilitySuperScraperFirecrawl
Clean, LLM-ready markdown
Content shaping (main-only / include-exclude tags)
Free JSON-LD / Open Graph extraction(where a site ships structured data)
Confidence score on extracted JSON
llms.txt generation

Operations

CapabilitySuperScraperFirecrawl
Automatic stealth-browser escalation
Proxy rotation handled for you
Per-domain cheapest-path cache
Never billed for failed scrapes

Distribution

CapabilitySuperScraperFirecrawl
Node + Python SDKs(in progress, not published)
CLI(in progress, not published)
Plain REST + bearer auth
MCP server
Compatible request shapes
Long-tail language SDKs (Go/Rust/Java/…)(on our backlog)

Performance

Measured 2026-06-23
MetricSuperScraperFirecrawl
Median scrape latency (p50)2.17swe have not measured this
Anti-bot pass rate, 27 live sites96.3% (26/27)we have not measured this
Paid-proxy spend on that run$0we have not measured this

Our numbers come from one dated run against 27 live sites on 2026-06-23. The script and the raw per-site rows are published, so you can re-run it: see the method and what it does not prove. The Firecrawl column is empty because we have not benchmarked their stack. We will not put a number there until we do.

Full support Partial Not available