PolySimulator HFT API v1
Build, test, and deploy prediction market trading bots in a risk-free environment powered by real-time Polymarket data. When you’re ready, switch to live trading by changing a single environment variable.
pip install → a 5-line script → a filled trade. Start in the Quickstart.
Quick Start
Get your API key and place your first trade in under 2 minutes.
Python SDK
pip install polysimulator — the official PolySimClient. Porting a py-clob-client or py-sdk bot? Drop-ins included.API Reference
Interactive playground — test every endpoint with your API key.
Example Bot
Copy-paste a working Python trading bot and start experimenting.
Why PolySimulator?
Architecture Overview
Your bot talks to a single endpoint. The API streams live order books, prices, and markets from Polymarket’s CLOB and Gamma feeds and serves them from an in-memory cache (Redis) — so your strategy reads current market data with low latency, without a database round-trip on the hot path. Orders execute against your simulated paper account, and the wire contract is identical to live Polymarket, so going live is a config change — not a rewrite.Base URL
Production:https://api.polysimulator.com
Endpoints are mounted under /v1 — append the version + path to the
base, e.g. GET https://api.polysimulator.com/v1/health. (The one
documented exception is /api/beta/cohort-status, a public read-only
endpoint used by the pricing page for closed-beta slot availability.)
Keep the base URL without /v1 so each path can prepend its own
/v1 — that’s the convention the Quick Start and
API Reference use.
One Config Change to Go Live
1
Build with PolySimulator
Develop and test your bot against the virtual API with a simulated API wallet
(25,000 on Pro+; Free-tier keys are read-only with no API wallet).
2
Validate Strategy
Review your portfolio, trade history, and equity curve to confirm your edge.
3
Switch to Live
Set
TRADING_MODE=live and add your Polymarket credentials. Same API, real money.AI Agent Integration (MCP)
PolySimulator docs are indexed for the most common AI-coding-tool integrations — a hosted MCP server,llms.txt, the machine-readable
OpenAPI spec, and Context7 — so Cursor, Windsurf, Claude Desktop,
Continue, and similar assistants get live access to the current API
schema. See Use with AI assistants for setup configs
and example prompts.
What’s Next?
Authentication
Learn how API key auth works and create your first key.
Place a Trade
Execute market and limit orders with string-precision numerics.
WebSocket Feeds
Subscribe to real-time price updates and execution notifications.
Error Handling
Handle rate limits, errors, and build retry strategies for your bot.