arcade-snaprender
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
arcade-snaprender
Arcade tools for SnapRender, so AI agents can see and read public web pages. Pages load in a real Chrome browser, so content drawn by JavaScript is included.
Tool | What it does |
| Captures a page as PNG, JPEG, WebP or PDF (full page, phone or tablet presets, dark mode, hidden elements, click before capture) and returns a link to the file |
| Reads a page as markdown, plain text, HTML, article fields, a list of links, or metadata |
| Shows the plan, captures used this month and captures left |
Built with arcade-mcp. Works as a local MCP server (stdio or HTTP) or deployed to Arcade Cloud.
Get an API key
Create a free account at app.snap-render.com. The free plan includes 200 captures a month with no card. Screenshots and extractions share the allowance; failed captures are not counted.
The tools read the key from an Arcade secret named SNAPRENDER_API_KEY. The key is never passed to the model.
Related MCP server: Screenshot API
Install
pip install arcade_snaprenderor with uv:
uv pip install arcade_snaprenderRun it
Local MCP server
Export the key, or put it in a .env file in the folder you start the server from:
echo "SNAPRENDER_API_KEY=sk_live_..." > .envThen start the server over stdio (for Claude Desktop, Cursor and other local clients):
arcade-snaprenderor over HTTP on 127.0.0.1:8000:
arcade-snaprender httpYou can also load it with the Arcade CLI, which finds installed arcade_* packages:
arcade mcp --tool-package snaprenderArcade Cloud
Clone this repository, log in and deploy:
arcade login
arcade deploy -e arcade_snaprender/__main__.pyThen add SNAPRENDER_API_KEY under Secrets in the Arcade dashboard, or with the CLI:
arcade secret set SNAPRENDER_API_KEY=sk_live_...Tools
TakeScreenshot
Parameter | Type | Default | Notes |
| string | required | Full URL starting with |
|
|
| |
| bool |
| Whole scrollable page |
|
| none | Overrides width and height |
| int |
| 320 to 3840 |
| int |
| 200 to 10000 |
| bool |
| Asks the page for its dark color scheme |
| bool |
| |
| bool |
| |
| string | none | Comma-separated CSS selectors to hide |
| string | none | Element to click before the capture |
| int |
| Extra wait in ms after load, 0 to 10000 |
| int |
| JPEG and WebP only |
| bool |
| Also return the file as a base64 data URI |
| int |
| Link lifetime in seconds, 60 to 2592000 |
Returns screenshot_url (a link anyone can open, ready to paste into Slack, an email or a ticket), link_expires_at, format, width, height, size_bytes, served_from_cache, render_time and remaining_credits.
Each capture uses one credit. When the same page with the same options was captured in the last 24 hours, that copy comes back for free. Opening the link serves the stored copy, so it does not use another credit while the copy is kept (24 hours on the free plan, longer on paid plans).
ExtractContent
Parameter | Type | Default | Notes |
| string | required | |
|
|
| |
| string | none | Read only part of the page, e.g. |
| bool |
| |
| bool |
| |
| int |
| 0 to 10000 ms |
| int |
| 1 to 500000 characters |
Returns url, type, content, word_count and processing_time_ms.
GetUsage
No parameters. Returns plan, period_start, period_end, used, limit and remaining. Free to call.
Limits
SnapRender only captures public pages. Pages behind a login, localhost and private network addresses are refused. If a site you own shows a security check instead of the page, see allowing SnapRender through your firewall.
Development
python -m venv .venv && . .venv/bin/activate
pip install -e ".[dev]"
pytest -q
ruff check . && mypyThe tests mock the SnapRender API, so they need no key.
Links
SnapRender docs: https://snap-render.com/docs
API reference (OpenAPI): https://app.snap-render.com/openapi.json
Pricing: https://snap-render.com/pricing
Support: support@snap-render.com
License
MIT
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
A real browser for your agent: screenshots, page checks, Markdown and PDFs.
Hosted browser for agents: open pages, search the web, extract data, read PDFs and take screenshots.
Read a URL as clean markdown, screenshot a website, url to PDF. Web access for agents, no signup.
Capture websites as PNG, JPEG, WebP or PDF, extract their content, or analyse them with AI.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables AI agents to capture clean web screenshots and extract main article content from any web page, with automatic removal of cookie banners, ads, and boilerplate.2MIT
- AlicenseNot gradedqualityDmaintenanceEnables AI agents to capture any public URL as PNG, JPEG, or PDF via REST API or MCP tools, including screenshot capture, page description, and PDF rendering.18 npmMIT
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to read, screenshot, or convert web pages to PDF using a real headless browser, turning any URL into clean Markdown, a visual image, or a print-ready document.82 npmMIT
- AlicenseNot gradedqualityBmaintenanceGives AI agents a real browser to capture screenshots, PDFs, and post-JavaScript HTML directly from URLs, with no API key or signup required.MIT