Most agent stacks eventually need the same unglamorous capabilities: resize an image, merge a PDF, screenshot a page, mint a UUID, format JSON, watch a public URL. Wiring a separate SaaS for each of those is a tax on every project.
Agent Media Tools is one free-to-start toolkit for humans and AI agents: browser UI, REST API, and MCP — same tools, same limits, one authentication model.
Try it in five minutes
The canonical path is /first-win: free UUID call → browser tool → free account → API key → authenticated status request. No credit card.
What you get
70+ tools across media, PDFs, web intelligence, developer utilities, and agent infrastructure
Free tier for anonymous and free-login use (higher daily limits after signup)
Pro at $5/month — shared 1,000 daily tool calls + 300 monthly media credits
MCP packages for Claude / Hermes-style hosts, plus OpenAPI and /llms.txt
Workflow playbooks for product images, document intake, and website monitoring
First success without an SDK install
Official PyPI/npm packages are reviewed but not published yet. Production integrations today use REST or MCP:
We do not advertise historical uptime we have not independently measured.
Private/reserved network destinations are blocked for SSRF safety.
Image OCR and some media paths depend on the installed host toolchain.
SDK install commands will appear only after registry publication.
Shareable copy
Show HN / short post
Show HN: 70+ media/PDF/web tools for AI agents behind one free API key
I got tired of gluing Cloudinary + PDF APIs + screenshot services for agents.
Agent Media Tools: browser UI + REST + MCP, free tier, Pro $5/mo.
First win: https://agentmediatools.com/first-win
Site: https://agentmediatools.com
X / Discord blurb
Give your agent hands, not just words.
70+ tools · free to start · $5 Pro · REST + MCP
→ https://agentmediatools.com/first-win