Render your HTML charts, reports, and pages to an image or PDF.
From a URL or Markdown too — one POST to a rendering API built for automation. Wire it into n8n, Make, Zapier, or any AI agent over MCP. Signed webhooks, batch, and 300+ edge locations.
Call → render → image. Three steps, no detour through a browser farm you have to manage.
POST your payload, get bytes back. Same engine across every plan, every region, every interface — REST, SDK, or MCP.
curl https://api.rendex.dev/v1/screenshot \
-H "Authorization: Bearer $KEY" \
-d '{"markdown":"# Invoice {{n}}","data":{"n":"A-100"},"format":"pdf"}'{ "engine": "chromium-124", "edge": "iad", "took_ms": 211 }No infra to provision. No browsers to crash-loop. We retire workers between requests so one bad page can't poison the next.
HTTP/1.1 200 OK Content-Type: image/png Content-Length: 84321 X-Rendex-Edge: iad · X-Rendex-Took-Ms: 211
n8n, Make, Zapier, MCP — the four ways most Rendex calls actually start.
Drop into an existing workflow without writing code, or call from any HTTP-aware tool.
n8n community node
Official n8n node with HTML, URL, and Markdown render modes — plus 10 ready-made workflow templates you can import in one click.
↳ DROP-IN STEPMake HTTP module
Pre-built HTTP scenario template — render HTML, a URL, or Markdown, then pipe the image or PDF into Drive, Slack, or Notion.
↳ TEMPLATEZapier action
Native Zapier action with rendering presets — social card, thumbnail, full-page capture, PDF report, OG image — from HTML, a URL, or Markdown.
↳ NATIVE ACTIONModel Context Protocol
One-line MCP endpoint so Claude, Cursor, and any MCP-aware agent can render images and PDFs as a tool call.
↳ TOOL ENDPOINTOne MCP endpoint — Claude, Cursor, and your own agents render in a single tool call.
Same render engine, exposed as MCP tools your agent runtime already understands.
Give your agent a render tool, not a browser farm.
MCP is how Claude, Cursor, and other agent runtimes load external tools. Rendex exposes its render engine as MCP tools — your agent gets render_artifact, rendex_screenshot, rendex_extract and more, instantly.
- Arender_artifact — Markdown or HTML → a branded PDF, PNG, and hosted share link in one call — built for agent outputs.
- Srendex_screenshot — HTML, Markdown, or URL → image or PDF. Same viewport, waitForSelector, and format options as REST.
- Erendex_extract — URL → clean reader-mode Markdown, JSON, or article HTML, pulled from the fully-rendered DOM.
- +render_link · account · watch — Mint hosted render links, check plan + usage in-chat, and monitor pages for changes — 13 MCP tools in all.
{
"mcpServers": {
"rendex": {
"url": "https://mcp.rendex.dev/mcp",
"headers": {
"Authorization": "Bearer $RENDEX_KEY"
}
}
}
}Seven things — picked because they matter, not because we needed to fill twenty-five tiles.
Render, PDF, Markdown & templating, hosted URLs, extraction, batch, reliability. Tab through what you care about.
HTML → image
Raw HTML in, PNG or JPEG out. ~95% of real Rendex traffic. Custom viewport, retina @2x, dark-mode emulation.
URL → image
Point at any URL. waitForSelector, networkidle wait, ad + cookie-banner blockers.
Full-page capture
Scroll-stitched screenshots up to 16,384 px tall.
Device presets
One device field — iphone_15, pixel_8, ipad — sets viewport, scale, and user-agent for you.
Resize & thumbnails
Downscale the output with resizeWidth / resizeHeight — aspect ratio preserved when you set just one — straight into an <img src> thumbnail.
{
"html": "<…>",
"format": "png",
"device": "iphone_15",
"waitForSelector": "img.hero",
"fullPage": true,
"blockCookieBanners": true,
"hideSelectors": [".chat-widget"],
"resizeWidth": 600
}Add a capability, watch it land three ways at once.
Point Rendex at a URL and toggle a capability. The result reshapes, the request grows a highlighted line for developers and AI agents, and the matching field lights up in a no-code node for n8n, Zapier, and Make. Same API, shown the way each person actually works.
Now it watches pages too, not just renders them.
Point Rendex Watch at any URL and get a real-Chrome before/after visual diff with a highlighted overlay — alerts by email (every plan) or webhook (Starter+). Create a watch point-and-click in your dashboard or with one API call, on the same rdx_ key and credit pool.

Four tiers. Hard cap on every plan. No overage, no asterisks.
Monthly or annual — 2 months free yearly. The cap is the deal, no surprise invoices, ever. EU VAT and B2B reverse-charge are captured at checkout.
Build, test, ship a prototype. Same engine, same edge — just a smaller monthly bucket.
- Calls / mo100
- Req / min3
- URLs / batch—
- Concurrent3
- API keys1
- Geo-target—
- Watches1 (daily)
- WatermarkTiled mark
Get off free without overpaying. Authenticated rendering and batch unlocked, 4× the free bucket — for evaluation and small side projects before you scale up to Starter.
- Calls / mo400
- Req / min10
- URLs / batch5
- Concurrent5
- API keys2
- Geo-target—
- Watches2 (every 3 hr)
- WatermarkNone
The full platform in production — every endpoint on one key. 20× the rate limit, 5 keys, bigger batches, and authenticated rendering.
- Calls / mo10,000
- Req / min60
- URLs / batch25
- Concurrent50
- API keys5
- Geo-target—
- Watches10 (hourly)
- WatermarkNone
For teams running production workflows. Geo-targeting, custom S3 storage and higher concurrency.
- Calls / mo100,000
- Req / min300
- URLs / batch100
- Concurrent200
- API keys10
- Geo-target50+ countries
- Watches50 (every 30 min)
- WatermarkNone
Higher limits, custom SLA, unlimited calls, and up to 1,000 website watches at 5-minute checks for teams running rendering at scale. We'll size a plan to your volume.
The questions we actually get from devs and automation builders.
Short answers, real code snippets, no marketing.
01How does the free tier work?
02Do free renders have a watermark?
03What image formats are supported?
04How fast are screenshots?
05Do you have official SDKs?
06Can I capture full-page screenshots?
07Can I capture a specific part of a page?
08Is there an MCP server for AI agents?
09What happens if I exceed my plan limit?
10Do you support dark mode captures?
11Where is the API deployed?
12Can I capture pages from different countries?
13How does batch processing work?
14Do you support webhooks?
15Can I generate PDFs from URLs or HTML?
16Can I render HTML to an image without hosting a page?
17Can I render Markdown to an image or PDF?
18Can I get an OG image as a hosted URL instead of bytes?
19Can I turn a URL into clean Markdown for an LLM or RAG pipeline?
20Can I generate invoices or reports from a template and data?
21Can Rendex monitor a page for changes?
22What's the difference between Rendex and a screenshot-only API?
23Is my data safe with Rendex?
24Do you scan URLs for malware before rendering?
25Can the API access internal or private URLs?
26How do I use Rendex in an n8n workflow?
27Can I render HTML to an image in Make.com?
28Does Rendex work with Zapier?
29Can I generate invoices or OG images from a no-code automation?
30Are you GDPR-compliant, and can I get a DPA?
31Do you provide a VAT-compliant invoice for my business?
Grab a key. Paste this. You're rendering.
Free tier is a real tier — same engine, same edge, same MCP. Lift the cap when you need to, not before.
curl "https://api.rendex.dev/v1/screenshot" \
-H "Authorization: Bearer $RENDEX_KEY" \
-H "Content-Type: application/json" \
-d '{"html":"<h1>Hello</h1>"}' \
-o out.png
