One API for AI agents to use the web: screenshot, HTML→PDF, read as Markdown, extract JSON, crawl, and monitor any URL. From $0.90 per 1,000. Drop-in for Urlbox/ScreenshotOne/Firecrawl.
Hi PH, I'm Amir, I built rasterly.
I kept hitting the same wall: every project needs to turn a webpage into an image or PDF (OG images, PDF invoices, previews, visual monitoring) and the existing APIs get expensive fast, or you end up running your own headless-Chrome fleet and babysitting it.
rasterly is that one primitive, cheap and boring in the good way:
- URL or raw HTML in, PNG / JPEG / PDF / GIF / MP4 out
- full-page, retina, wait-for-selector, transparent bg, direct S3 upload
- SSRF-guarded, flat pricing from $0.58 per 1,000 (up to ~80% under the incumbents)
- drop-in: swap the host + key and you're done, I'll migrate you for free
Two things the incumbents don't have:
- an MCP server, so an AI agent can screenshot / make a PDF as a native tool
- x402 pay-per-call, a wallet pays per request, no plan, no signup
Free-tier key is instant (email in, key out). Would love feedback on the API design and what's missing, reply here and I'll answer all day.