Institutional

The Fed record, wired into your stack

Everything in Pro, plus the REST API, an MCP server for your AI tooling, the raw intraday tone feed, and seats for the desk.

$499/ monthor $4,990 / year
Statement push
< 30 s (p95)
Whisperer alerts
< 60 s from publication
Missed FOMC releases
Zero — the standard we engineer to
API

Four export endpoints, one envelope

Read-only, cursor-paginated exports of the same derived data behind the app. Every request takes a time window (90 days by default, one year at most) and every response carries the compliance disclaimer in the payload. Derived and aggregated values only — we never redistribute raw licensed market data.

GET /api/v1/data/tone-scores

Document-level dovish–hawkish scores across the communication record.

id, document_id, published_at, score, method
GET /api/v1/data/statement-diffs

Every statement diff with its material-change count and net direction.

id, new_doc_id, old_doc_id, published_at, material_changes, net_direction, diff_json
GET /api/v1/data/divergence

The daily market-vs-Fed divergence series with percentile context.

as_of, divergence_bps, percentile, market_path, fed_path
GET /api/v1/data/playbooks

Full playbook statistics per language pattern, including the n ≥ 3 honesty flag.

pattern_key, display_name, action, window, n, insufficient_history, medians, iqr
curl "https://api.fedreact.com/api/v1/data/tone-scores?start=2026-06-01T00:00:00Z" \
  -H "Authorization: Bearer $FRE_API_KEY"

{
  "items": [
    {
      "id": 4182,
      "document_id": 913,
      "published_at": "2026-06-17T18:30:00+00:00",
      "score": 24.0,
      "method": "lexicon_v1"
    }
  ],
  "next_cursor": "NDE4Mg==",
  "window": { "start": "2026-06-01T00:00:00+00:00", "end": "2026-06-30T00:00:00+00:00" },
  "disclaimer": "Historical data for informational purposes only. Not investment advice. Past performance does not predict future results."
}

Sample shown for illustration; live keys are provisioned per organisation.

MCP

Fed intelligence inside your AI assistant

The MCP server wraps the institutional API one-to-one — same key, same rate limits, same metering, so it is never a side door around the API quota. It works with Claude Desktop, Claude Code, and any MCP-compatible client.

get_fed_calendar

Upcoming Fed events and blackout windows.

get_statement_diff

A meeting's statement diff with materiality and net direction.

get_pattern_playbook

Medians, IQR, occurrence count, and instance-level reactions for a language pattern.

get_tone_scores

Document and sentence tone scores with the key phrases behind them.

get_speaker_stats

Speaker impact ranking and market-moving rates.

get_divergence

The latest divergence reading, its percentile, and the series.

Every tool response carries the disclaimer and a one-line summary, so when your assistant quotes the data inside a conversation, the attribution and the compliance framing travel with it.

Reliability
< 30 s

statement release to push notification, p95

< 5 s

statement ingestion to computed diff

< 3 s

spoken word to live tone score during pressers

Zero missed releases

every release-day path has heartbeat alerts and a rehearsed dry-run runbook; missing an FOMC release is treated as an existential failure, and we build accordingly.

Every plan, compared
FeatureFreeProInstitutional
Calendar + countdowns
Statement diff20-min delayReal-timeReal-time
PlaybooksMedians only, 3/monthFull + instance dataFull
Whisperer alerts
Live tone scoringEnd-of-event summaryReal-timeReal-time + raw feed
Divergence gaugeWeekly readingDaily + alertsIntraday
Speaker analyticsTop 3FullFull
Portfolio exposure + scenarios✓ (25 positions)✓ (unlimited)
API access✓ (rate-limited)
MCP server✓ (same limits as API)
Seats115

Questions about billing or trials? See the FAQ.

Who runs it

Macro PM

Meeting prep in one screen: the diff history, the playbook for each pattern in play, and where market pricing sits against the dots. Then a per-scenario estimated sensitivity table for the book, built on published betas — no black box.

Systematic desk

Pull tone scores, statement diffs, and playbook statistics into your research pipeline as windowed, cursor-paginated JSON. Every score records its method and methods are versioned, so backtests stay reproducible.

AI-native fund

Point your internal assistant at the MCP server and analysts can query the Fed communication record mid-conversation — structured responses with a quotable summary, attribution, and the disclaimer attached to every payload.

Talk to us about a rollout

Tell us about your desk and how you want to consume the data — dashboards, research pipeline, or assistant. We will set up seats, keys, and the feed, and share the scoring validation report and uptime record.

Historical data for informational purposes only. Not investment advice. Past performance does not predict future results.