taskminer-render

Paid rendering API for AI agents - turn text into MP4 video or PNG images via x402 payments on Base

Documentation

taskminer-render

Paid rendering API for AI agents. Turn a text brief into a real MP4 motion graphic or PNG image. Built for agent-to-agent commerce on x402.

curl -X POST https://taskminer.exe.xyz/v1/render/image \
  -H "Content-Type: application/json" \
  -d '{"brief": "a bold geometric poster reading HELLO WORLD in neon green on black"}'
-> 402 Payment Required (0.25 USDC on Base)

Pricing

EndpointPriceOutput
POST /v1/render/image$0.25 USDC1600×900 PNG (or custom size)
POST /v1/render/video$0.75 USDC1920×1080 MP4 H.264 (3–30 seconds)

Payment is x402 v2 on Base mainnet (USDC). No signup, no API key — pay per call, receive the artifact in the response.

MCP (paid tool calls for agents)

MCP endpoint: https://taskminer.exe.xyz/mcp/
Transport: streamable HTTP

Two tools: render_image and render_video. tools/list is free. tools/call is paid — same prices as HTTP. Clients using @x402/mcp or Cloudflare Agents SDK auto-pay.

How it works

  1. Your agent sends a brief
  2. An LLM (DeepSeek-V4-Pro) converts it into an SVG or HyperFrames HTML composition
  3. The composition is rendered to a real file in a locked-down container (Chromium/rsvg)
  4. The artifact is returned

No uploads, no queue, no human review. Fully deterministic: same brief = same output.

Distribution / discoverability

  • x402scan: registered at https://www.x402scan.com
  • MCP directories: Smithery, mcp.so, mcpservers.org, glama.ai (auto-crawled from this repo)
  • Official MCP registry: pending submission

Deploy your own

git clone https://github.com/4nkitd/taskminer-render
# requires: Docker, Node 22, ffmpeg, librsvg2
# see x402/ directory for the FastAPI + FastMCP server

Security

Buyer briefs execute inside two hardened Docker containers (no network, no credentials, read-only rootfs, non-root user, dropped capabilities). The renderer can never reach the wallet.


Built for taskmarket.dev and the x402 payments ecosystem.