Vike Data API docs: on-chain, Hyperliquid, Polymarket

On-chain, Hyperliquid, HL Cohorts and Polymarket data — over REST, MCP, and bulk Parquet files.

The Vike Data API serves on-chain and market data in four products — On-chain, Hyperliquid, HL Cohorts and Polymarket — as JSON over REST, as tools for AI agents over MCP, and as bulk Parquet files for research-scale downloads. It is free to start.

Pick a product

On-chain

Wallets, tokens, transfers and labels on Ethereum, BNB Chain, Base and HyperEVM.

Hyperliquid

Every wallet's perp history, positions, liquidations, builders, candles and daily Parquet files.

HL Cohorts

How whales, small accounts and profitable traders are positioned, per coin and per hour.

Polymarket

Markets, the order book and its archive, candles, trader profiles, smart money and settlements.

Cross-product

Questions no single product answers, with one key across all four.

Each product page starts with the questions it answers, as examples: one question, a call you can copy and a real response. The datasets behind it, its endpoints and what to know before you rely on it follow.

Ways in

Way in	Where	Auth
REST	`https://data.vike.io/v1`	`X-API-Key` header
MCP	`https://data.vike.io/mcp`	connect from Claude or any MCP client
CLI	`vike-data` (npm `@vike-io/cli`)	the same key, saved by `vike-data init`
Bulk archive	`https://data.vike.io/archive`	`X-API-Key` for Polymarket data files; the Polymarket `manifest.json` and the free samples of both archives are open. A Hyperliquid file is a 7-day link from `/v1/hyperliquid/archive/url`

REST and MCP share one hourly budget per API key (your OAuth tokens share one between them); archive downloads are not charged against it — see Limits and errors.

REST is plain JSON over HTTPS across the 106 operations of the /v1 spec. Getting started makes the first call, and REST conventions holds the rules that apply to every namespace.

MCP puts the same data behind tools an agent calls conversationally. Claude Desktop connects over OAuth with no API key at all. All 73 tools are documented under Build with AI, and Client setup walks each client.

The CLI, vike-data, puts those tools in a terminal, reads the live event streams and ships the agent skills; Data CLI covers it.

The bulk files come in two archives, both Parquet, and neither download spends the hourly budget:

Polymarket is a date-partitioned tree with Hive-style paths under https://data.vike.io/archive/. Its manifest.json and the free daily samples under /archive/samples/ are open; a data file needs your key. Parquet archive describes it and Downloading shows how to fetch it.
Hyperliquid is one file per UTC day of fills and of 5-minute position snapshots, plus a latest positions file, but there is no folder: you ask /v1/hyperliquid/archive/url for a link to one file, valid 7 days and served by Cloudflare R2, and the link itself is the credential. Two small free samples are open without a key under /archive/hl-samples/. Download the Hyperliquid file archive shows how.