Any crypto market → a token-compact brief, in one call.

What you get

One call in, the whole market state out

POST https://patternfetch.com/v1/brief
Authorization: Bearer pf_…

{ "ticker": "BTC/USDT",
  "timeframe": "4h" }

{
  "header": { "sym":"BTC/USDT","tf":"4h","src":"binance","n":200 },
  "codec": { "rows":"60125.4,60480,59890.1,60310.7,1284;…",
             "sax":"dcefdcbe", "precision":1 },
  "analysis": {
    "patterns":[{"name":"double_bottom","confidence":0.86}],
    "levels":{ "support":[{"price":59820.4,"strength":1}],
               "resistance":[{"price":63450.8,"strength":1}] },
    "regime":{ "trend":"up","strength":0.42,"volPct":2.13 },
    "indicators":{ "rsi":{"v":58.3,"state":"neutral"},
                   "ema":{"v":61240.8,"state":"above_20_50"} },
    "nl":"BTC/USDT: uptrend (moderate), +1.94% last 4h,
          RSI 58.3 (neutral), double_bottom (conf 0.86)."
  }
}

The whole picture in a few hundred tokens instead of a multi-thousand-token candle array — and the nl line is a ready-to-reason summary, so the model never has to do arithmetic on raw numbers (where LLMs hallucinate). See it live →

Quickstart (REST)

0. Try it with no signup — a real brief, no key:

curl -X POST https://patternfetch.com/v1/demo \
  -H "Content-Type: application/json" \
  -d '{ "ticker": "BTC/USDT", "timeframe": "4h" }'

1. Get a free key (includes $0.05 starter credit):

curl -X POST https://patternfetch.com/v1/keys \
  -H "Content-Type: application/json" \
  -d '{ "email": "you@example.com" }'
# -> { "key": "pf_…" }   (shown once — store it)

2. Get a brief for a ticker + timeframe:

curl -X POST https://patternfetch.com/v1/brief \
  -H "Authorization: Bearer pf_…" \
  -H "Content-Type: application/json" \
  -d '{ "ticker": "BTC/USDT", "timeframe": "4h" }'

Want only what changed since last poll? Call /v1/delta — it returns changed:false (billed at a fraction) when nothing material moved. Full reference in the docs.

Endpoints

Machine-readable: OpenAPI spec · llms.txt · methodology: how the signals are computed

For AI agents (MCP) — connect in one click

patternfetch ships an MCP server (Streamable HTTP) with four tools: patternfetch_brief, patternfetch_delta, patternfetch_analogs, and patternfetch_capabilities. Two ways to connect, both zero-friction:

A) OAuth — no key handling (Smithery, Claude.ai, Cursor, Claude Desktop)

Add the server URL and click Authorize once. patternfetch mints a free-tier key for you automatically — nothing to paste.

# Claude.ai → Settings → Connectors → Add custom connector
https://patternfetch.com/mcp
# (discovery: /.well-known/oauth-protected-resource)

B) Bearer key (config file)

{
  "mcpServers": {
    "patternfetch": {
      "url": "https://patternfetch.com/mcp",
      "headers": { "Authorization": "Bearer pf_…" }
    }
  }
}

Or one line: claude mcp add --transport http patternfetch https://patternfetch.com/mcp --header "Authorization: Bearer pf_…"

Pay without an account (x402) — agents top up credits with USDC over /v1/topup, no signup or human in the loop. Discoverable via /.well-known/x402.json.

Pricing

Pay per call — brief $0.01, candles $0.005, delta $0.008 (empty $0.001), analogs $0.05. A free tier covers trying it out and light use. Pay with a Stripe card or x402 (USDC) with no account. Full breakdown on the pricing page.

Built for

One /v1/brief call answers all of these — explore the detail:

Why patternfetch

Raw OHLCV is the wrong shape for an LLM: it's token-expensive and invites numeric hallucination. patternfetch returns the digested market state — patterns, levels, regime and indicator state — plus a one-line summary an agent can act on. One clean schema, crypto spot, minute-to-daily, with an MCP server and pay-per-call billing that an agent can settle on its own.

FAQ