Looking for a Firecrawl alternative for turning web pages into Markdown? Here's an honest comparison. Firecrawl is a full web data platform (scrape, crawl, map, search, structured extraction) sold as monthly plans with a recurring free tier. Shotmint does one part of that: GET /v1/content loads a single URL in headless Chromium and returns its main content as Markdown, text or clean HTML, paid per page with prepaid credits and no monthly plan. The same credits also buy screenshots and PDFs. For single pages it is compatible with Firecrawl's scrape call, so the Firecrawl SDKs work by changing the base URL and key.
| Model | Price | Free tier | Unused volume | |
|---|---|---|---|---|
| Firecrawl | Monthly or annual subscription | Hobby $16/mo for 5,000 pages · Standard $83/mo for 100,000 · Growth $333/mo for 500,000 (billed yearly; Hobby is $19 billed monthly). Extra credits from $5 per 1,000 | 1,000 pages a month, no card | Monthly allowance; rollover only on Scale (one month) |
| Shotmint | Prepaid credits, no subscription | €5 / 2,000 · €19 / 10,000 · €59 / 50,000 (one-time) | 25 free renders a month, no signup | Never expire |
Firecrawl prices are in USD and come from their public pricing page on 22 October 2026. Prices change, so check their current pricing. Shotmint prices are in EUR.
Firecrawl does far more than one page at a time: it crawls whole sites, maps their URLs, runs web searches, extracts structured JSON with an LLM, and drives a browser for interactive pages, with SDKs and integrations for most AI frameworks. It is open source (AGPL-3.0) if you want to self-host. Its free plan renews 1,000 pages every month, which covers small projects for free, and at high steady volume its Standard plan (100,000 pages for $83 a month, billed yearly) is cheaper per page than Shotmint. If you need crawling, search or extraction, or you scrape more than roughly 60,000 pages every month, Firecrawl is the better choice.
You already know which URLs you want and need them as clean Markdown, your volume is between a few thousand and a few tens of thousands of pages a month, or it is irregular. 5,000 pages a month cost $192 a year on Hobby (billed yearly) against €78 a year on Shotmint (one €59 Scale pack plus one €19 pack), and 20,000 a month cost $996 a year on Standard against €295 (five Scale packs). Credits never expire, so a quiet month costs nothing, and a failed render never costs a credit. /v1/content returns title, author, date, language and word count with the content, can drop link URLs and images (links=false) to save tokens, and takes cookies and headers for pages behind a login. PDF links work too: the same call returns the text of a PDF as Markdown, with headings and links. See the web page to Markdown for LLMs guide for measured token counts and RAG code. AI agents can also call it as the read_page tool on our MCP server.
We serve Firecrawl's POST /v2/scrape and /v1/scrape, so existing scrape code keeps running:
from firecrawl import Firecrawl
fc = Firecrawl(api_key="YOUR_KEY", api_url="https://shotmint.creasework.com")
doc = fc.scrape("https://example.com", formats=["markdown"])
print(doc.markdown)We ran this on 5 November 2026 with firecrawl-py 4.55.0, and the same call with @mendable/firecrawl-js 4.52.0 (apiUrl). Formats: markdown, html, rawHtml and links, 1 credit per call whichever you ask for. Also read: onlyMainContent, waitFor, mobile, blockAds, headers. Crawl, map, search, extract, batch, screenshot and JSON formats, actions, includeTags and excludeTags return an error that says so rather than a partial result. The Markdown comes from our own extractor, so it will not match Firecrawl's byte for byte: compare a few of your pages on the free tier first.
GET /v1/screenshot: PNG, JPEG or WebP screenshots of any URL, with full-page capture, custom viewport, retina scale, a single-element selector, dark mode, transparent backgrounds (omit_background), ad and cookie-banner blocking (block_ads), and hide_selectors.POST /v1/pdf: PDFs from a URL or raw HTML, with format, landscape, margins, backgrounds, header and footer templates, and page numbers.GET /v1/content: the main content of any page or PDF link as Markdown, plain text or clean HTML, with title, author and date, for LLMs and RAG.GET /v1/og: 1200×630 Open Graph cards from a title and subtitle, with signed public URLs you can put straight in an og:image tag.X-Credits-Remaining header.Shotmint is a new, small service. It has small Node and Python SDKs and integration guides, but no enterprise plans, SLA or team accounts yet (measured uptime and render times are public on /status). If you need those, Firecrawl may suit you better. If you want an API key in 10 seconds and to pay only for what you use, try it.
ScreenshotOne, Urlbox, PDFShift, ApiFlash, PagePixels, ScreenshotAPI.net, HTML/CSS to Image, Microlink, Screenshot Machine, Screenshotlayer, Bannerbear, Placid, Browshot, Restpack, ScrapingBee, DocRaptor, pdflayer, html2pdf.app, Browserless, PDFMonkey, Doppio, CaptureKit, Pdfcrowd, APITemplate.io, url2png, CraftMyPDF, IronPDF, SelectPdf, Prince, Aspose.HTML, EvoPdf, PDF.co, CloudConvert, pdfEndpoint, Abstract Screenshot API, GrabzIt. Yearly cost at three volumes: screenshot API pricing and HTML to PDF API pricing.
No signup, no card; the free key refills on the 1st of every month. If you need more, buy credits once: €5 for 2,000, €19 for 10,000, €59 for 50,000. They never expire, and they stack on top of the 25 free renders a month, which every key keeps. From a script, an AI agent, Make or Zapier, POST /v1/keys/free (no body) returns a key as JSON (openapi.json, no-code recipes).