Sports Intelligence API
Structured, real-time sports intelligence with rationale, confidence scoring, and change attribution — purpose-built for autonomous systems that reason and act.
No credit card required · 1,000 free credits to start · REST + MCP
// Agent invocation
GET /v1/events/9199/intelligence
// Structured intelligence response
{
"event_id": 9199,
"available": true,
"has_recommend": true,
"analyst_take": "Sharp money has moved toward San Diego through the morning.",
"bets": [{
"bet_type": "ML_P1",
"player_name": "San Diego Padres",
"tier": "moderate",
"confidence_score": 0.611,
"rationale": ["Strong Starting Pitching Edge (20/25)", "Deep Research supports this bet (+3.1pp)"],
"attribution": ["pitching_edge", "deep_research"]
}]
}Lumify is a real-time sports intelligence API built for AI agents. It turns fragmented sports data into structured, explainable signals — confidence scores, rationale, and change attribution — that autonomous systems can reason over and act on directly, instead of raw feeds built for human dashboards.
They were built for dashboards. Lumify is built for systems that reason, decide, and act.
Fragmented sportsbook events, odds, and player data normalized into a single, deterministic schema. One ID. Every source.
Odds and market shifts are captured with full movement history and
moved_at
timestamps via
/odds/history,
so agents can detect changes without re-deriving state.
Every output carries a
rationale array,
confidence_score,
and attribution.
Not just what — but why, how confident, and from where.
RESTful endpoints with deterministic schemas, OpenAPI docs, and structured outputs ready for agent reasoning loops. One API key. Every major sport.
# 1. REST-only — use any HTTP client
import requests
# 2. Authenticate with your Bearer key (lmfy-...)
headers = {"Authorization": "Bearer lmfy-..."}
# 3. Query structured intelligence for an event
signal = requests.get(
"https://lumify.ai/v1/events/9199/intelligence",
headers=headers,
).json()
# Agent can reason over structured output
if signal["has_recommend"]:
# bets[] isn't sorted by confidence — pick the first non-"avoid" tier
top = next(b for b in signal["bets"] if b["tier"] != "avoid")
agent.act(top["rationale"], top["confidence_score"])Agents that monitor line movement, detect sharp money, and execute decisions in real time — powered by structured signals with embedded rationale.
Quantitative systems that need canonical cross-sportsbook mapping and volatility signals without building their own ingestion infrastructure.
Automated insight generators, prediction pipelines, and content tools that require live, structured intelligence — not raw data feeds to parse.
A real-time sports intelligence API designed for AI agents and autonomous systems. It delivers structured, explainable data — including live scores, player signals, market intelligence, and change attribution — in a format agents can reason over and act on without additional processing.
NFL, NBA, MLB, NHL, NCAA Football, NCAA Basketball, ATP/WTA tennis, and the FIFA World Cup for soccer, with schedules, live scores, and odds across all of them. Full bet intelligence is live today for MLB, NFL, NCAA Football, tennis, and FIFA World Cup soccer.
Both. npm install @lumifyai/sdk
gives you an official, typed TypeScript/JavaScript client, and the REST API
works with any HTTP client. A hosted MCP server is also available for
Cursor, Claude, and other agent tooling.
1,000 credits valid for 30 days, full access to every endpoint, up to 2 active API keys, and a 10 requests/minute rate limit — no credit card required to start.