Firecrawl turns a page into clean markdown. webreadr hands you the raw material behind it: the post-JS HTML verbatim, the structured data the page already ships, and the XHR/fetch responses behind single-page apps — over a self-hosted escalating anti-bot ladder. A near drop-in API for when markdown isn't enough.
Free tier: 5,000 credits/month · no card required · Firecrawl-shaped API.
Both clear bot walls and return LLM-ready data. The real difference is how much of the page you get back — and whether the data trapped inside a SPA ever reaches you.
| Capability | webreadr | Firecrawl |
|---|---|---|
| Raw post-JS HTML returned verbatim (never transformed) | ✓ Yes | ~ Partial |
| Structured data the page already ships (JSON-LD, OpenGraph, embedded state) | ✓ Yes | ~ Partial |
| Captured XHR / fetch responses behind SPAs (the network format) | ✓ Yes | ✗ No |
| Clean LLM-ready markdown | ✓ Yes | ✓ Yes |
| Self-hosted escalating anti-bot ladder (cheapest level first) | 6 levels | ~ Partial |
| Per-domain level caching (next request skips the climb) | ✓ Yes | ✗ No |
| Scrape / Crawl / Map / Extract endpoints | ✓ Yes | ✓ Yes |
| Firecrawl-shaped API surface (near drop-in) | ✓ Yes | — |
| Schema / prompt structured extraction | ✓ Yes | ✓ Yes |
| Credit-based pricing (pay for what a request costs to serve) | ✓ Yes | ✓ Yes |
| Free tier without a card | 5,000 credits / mo | 1,000 credits / mo |
| Paid tiers live today | ✗ No | ✓ Yes |
| Search-the-web endpoint | ✗ No | ✓ Yes |
| Mature SDK + integration ecosystem | ~ Partial | ✓ Yes |
Firecrawl pricing and feature points reflect its public plans (Free 1,000 credits/mo; paid tiers from $16/mo). “Partial” means the capability exists in some form but isn't the default or first-class output. We update this as both products ship.
We'd rather you choose well than choose us. Here's where each one genuinely shines.
The API surface is intentionally Firecrawl-shaped. Point the host at webreadr, swap the key, and add the formats Firecrawl never hands back.
# Firecrawl
curl https://api.firecrawl.dev/v1/scrape \
-H "Authorization: Bearer fc-***" \
-H "Content-Type: application/json" \
-d '{ "url": "https://example.com", "formats": ["markdown"] }'
# webreadr — same shape, point the host at us and swap the key.
# Ask for the raw material Firecrawl doesn't hand back:
curl https://webreadr.com/v1/scrape \
-H "Authorization: Bearer wr_***" \
-H "Content-Type: application/json" \
-d '{
"url": "https://example.com",
"formats": ["rawHtml", "structured", "network", "markdown"]
}'See the full request/response shape in the scrape docs.
Close. webreadr deliberately mirrors Firecrawl's API shape — the same scrape, crawl, map and extract endpoints, the same JSON request/response style and a bearer token — so for most scrape and crawl calls you point your base URL at webreadr.com, swap the key, and you're running. The main difference is what comes back: webreadr returns raw HTML, structured data and network responses by default, where Firecrawl centres on markdown.
When you need the data the page actually ships, not a readability summary of it. webreadr returns the post-JS HTML verbatim, the JSON-LD / OpenGraph / embedded state the page already carries, and — uniquely — the XHR/fetch responses captured during render, which is the only reliable way to get prices and content out of true single-page apps. If your pipeline only needs clean article markdown, Firecrawl is a fine choice too.
webreadr runs a six-level self-hosted ladder, from a fast plain fetch up to an anti-detect browser on a residential proxy, with a paid cloud engine as the last resort. In auto mode it starts cheap and only climbs when a site actually blocks it, then pins the winning level per domain so the next request skips the climb. You pay for speed on the easy 90% and only spend the heavy levels on the sites that fight.
Both bill on credits now, and webreadr's free tier is more generous (5,000 credits/month vs Firecrawl's 1,000, no card on either). Like Firecrawl, a webreadr credit's cost scales with the work — a cheap level-1 scrape is 1 credit, escalating the ladder or pulling premium formats costs more. Paid billing is not live on webreadr yet — Firecrawl has real paid tiers today (from $16/mo). If you need to scale to paid volume right now, that's a point in Firecrawl's favour; we're being upfront about it.
Billing and high-volume paid plans are live today, the SDK and integration ecosystem is more mature, and it ships a web-search endpoint webreadr doesn't have yet. If those matter more to you than raw-HTML and network capture, Firecrawl may be the better fit — and that's an honest answer.
Sign up, issue a key, and run the same URL through both. The difference is in the network tab.