NaxosDataNaxosData

Synchronized L2 order-book and tick data across prediction markets.

We record every millisecond WebSocket delta and deterministic Level 2 book state across Polymarket, Kalshi, ForecastEx, Limitless, PredictIt, Drift BET, and Smarkets—normalizing reciprocal binary books into high-performance archives and real-time streams.

6 Prediction Venues:Polymarket·Kalshi·Limitless·PredictIt·Drift BET·Smarkets
Live L2 Matched Pairs
pair_id synchronized
Fed Rate Cut Nov 2026Macro & Fed
Canonical ID: PAIR_FED_NOV26_25BPS
VenueBest BidBest Ask24h Vol
Polymarket
$0.635$0.640$4.2M
Kalshi
$0.618$0.622$2.8M
Limitless
$0.620$0.645$620K
PredictIt
$0.610$0.630$340K
NET ARB SPREAD
+1.8¢(180 bps)
SETTLEMENT FUNGIBILITY
98/ 100High Equiv
Open market in Terminal Inspector

Structural Microstructure Challenges

Prediction markets are forward-only, mathematically inverted, and non-fungible.

Standard equity and crypto trading infrastructure breaks when applied to prediction markets. We engineer the normalization layer so quantitative algorithms treat prediction markets as standard Level 2 order books.

01. DATA RETENTION

No backfill APIs
(Forward-only depth)

Prediction market exchanges only serve historical trades or coarse 1-hour candles. True resting Level 2 order book depth, queue cancellations, and historical bid-ask spreads vanish permanently unless recorded continuously in real time.

Solution: Continuous tick ingestion + Millisecond L2 Reconstruction.
02. MICROSTRUCTURE

Incompatible market
microstructure

Kalshi returns reciprocal binary YES/NO bids where Ask_YES = 1.00 - Bid_NO, while Polymarket uses conditional tokens and NegRisk multi-outcome books. We mathematically invert reciprocal bids into canonical USD bid/ask ladders.

Solution: Real-time reciprocal normalization into equities-style L2.
03. BASIS RISK

66% settlement
non-fungibility

Contracts that appear identical across Polymarket and Kalshi frequently diverge on oracle source (UMA vs. CFTC/AP), timezone cutoff (11:59 PM ET vs. midnight UTC), or resolution rules. Every matched pair_id in NaxosData includes explicit settlement provenance.

Solution: 0–100 Fungibility Scoring & settlement oracle catalog.

01. Technical Architecture & Ingestion

Capture, normalize, score and backtest.

Four institutional capabilities designed for quantitative desks, arbitrage bot operators, and market makers.

Un-Backfillable Historical L2 Archive

Partitioned ZSTD Parquet tick depth across 7 venues with deterministic millisecond Level 2 order book reconstruction.

View archive schemas

Canonical Symbology & Fungibility Scoring

0–100 oracle alignment, settlement rule verification, and automated cluster mapping across political, macro, and sports contracts.

View symbology mapping

Asymmetric Net-of-Fee Engine

Quadratic variance taker curves, 10% profit taxes, and maker/taker commissions normalized across all 6 prediction venues.

Explore fee models

Local SDK & 5-Strategy Backtest Lab

Thin-client Python SDK and discrete event simulation engine with safe AST custom strategy evaluation and latency modeling.

Launch Backtest Lab

03. Quantitative Workflows

Engineered for systematic desks and algorithmic traders.

Cross-Venue Arbitrage

Executing simultaneous cross-venue orders across Polymarket, Kalshi, ForecastEx, Limitless, PredictIt, Drift BET, and Smarkets, modeling latency slippage and unhedged sweeps.

Execution models →

HFT Market Makers

Modeling queue priority, adverse selection risk, and inventory skew on 5m/15m/1h crypto strike ladders and fast-breaking macro markets.

Queue priority models →

Macro & Event Desks

Extracting implied probability term structures around FOMC meetings, CPI releases, nonfarm payrolls, and geopolitical elections.

Term structure models →

Quant & Academic Research

Querying extensive historical tick and order book datasets directly via the NaxosData Historical Archive and Python SDK.

Research query snippets →

04. Transparent Self-Serve Pricing

Instant API provisioning. Flat-rate monthly subscriptions.

No opaque sales calls or surprise bandwidth surcharges. Choose your tier, receive an immediate API key, and begin querying historical tick archives.

Hobbyist
$0/ month

For exploratory researchers and retail prediction market quants.

Rate Limit
60 req/min
Lookback
3-day lookback
Export Quota
500k rows/mo
  • Standardized 20-level L2 depth snapshots
  • Live WebSocket orderbook delta feeds
  • Official Python & TypeScript SDK access
  • Claude Desktop & Cursor IDE MCP tools
  • Deterministic local paper backtester
Start Free
Quant ProMost Popular
$30/ month

The industry standard for active arbitrage operators & bot engineers.

Rate Limit
360 req/min
Lookback
90d archives
Export Quota
15M rows/mo
  • Real-time cross-venue arbitrage pair scanner
  • Settlement Fungibility Scores (0–100) & rule mismatch flags
  • Pre-trade slippage & execution impact simulator
  • Hive-partitioned Parquet (ZSTD) tick downloads
  • Multi-venue delta rebalancing & order routing
Subscribe to Quant Pro
Institutional
$100/ month

Unrestricted tick-level infrastructure for macro funds, market makers & primes.

Rate Limit
1,800 req/min
Lookback
Full archive
Export Quota
100M rows/mo
  • Sub-millisecond tick reconstruction & DuckDB views
  • Dedicated high-priority execution endpoints
  • Real-time Telegram & Discord webhook alert streams
  • Custom canonical pair mapping & symbology rules
  • 99.9% uptime SLA & quantitative engineer support
Get Institutional Access

Instant Deployment

Start querying prediction market tick data now.

Create an account in 30 seconds. Access our interactive no-code market explorer on the free tier ($0/mo), or subscribe to Quant Pro ($30/mo) for complete millisecond tick files.