Convert Social Posts to Markdown — an API Built for AI Agents

One POST request turns a Twitter/X thread, LinkedIn post, Medium/Substack/Beehiiv article, or public Notion page into clean, LLM-ready markdown. Pay $0.01 USDC per call over the x402 protocol — no API key, no signup, no subscription. Your agent discovers the price, pays, and gets markdown back in a single request cycle.

Endpoint

POST https://api.posttosource.com/api/x402/extract

Request body: {"url": "<link to the post>"}. An unpaid request returns 402 Payment Required with machine-readable payment options (x402 v2, USDC on Base). A paid request returns:

{
  "markdown":  "# ...",
  "title":     "...",
  "author":    "...",
  "wordCount": 123,
  "sourceUrl": "...",
  "platform":  "twitter | linkedin | medium | substack | beehiiv | notion"
}

Try it from the command line

With an x402-enabled client (for example awal, Coinbase's Agentic Wallet CLI) and a wallet holding USDC on Base:

npx awal x402 pay https://api.posttosource.com/api/x402/extract \
  -X POST \
  -h '{"Content-Type":"application/json"}' \
  -d '{"url":"https://x.com/naval/status/1002103360646823936"}' \
  --max-amount 10000

Discovery

OpenAPI specification: api.posttosource.com/openapi.json (includes x-payment-info and full request/response schemas)
x402 manifest: api.posttosource.com/.well-known/x402
Free capability check: GET https://api.posttosource.com/api/x402/extract returns this resource's description and input schema, no payment required
Also listed in the Coinbase CDP x402 Bazaar and on x402scan

Pricing and payment

$0.01 USDC per successful call — flat, every platform
Settlement on Base mainnet (eip155:8453) via the x402 exact scheme; gasless for the payer
Failed extractions are not charged — payment settles only after a successful response

Intended use

This API converts individual, publicly accessible pages that the requester is authorized to view — for personal reference, note-taking, and interoperability. It is not for bulk archiving, mirroring, systematic reproduction of websites, or building ML-training datasets. See our Terms of Service.

Prefer a UI?

Humans get the same conversion with a paste-a-link interface, PDF output, and NotebookLM-ready formatting on the PostToSource homepage.