Prediction Markets MCP Tool Reference

The Prediction Markets MCP server currently exposes 14 tools.

Exchange and market discovery tools

exchanges_list

Returns all supported prediction market exchanges.

Arguments

This tool does not require any arguments.

Returns

Each item can include fields such as:

• exchange_id
• market_name_institution_description
• website

Use when

• You need to discover supported venues before making market-level requests.
• You want the canonical exchange_id values exposed by the MCP server.

exchanges_get

Returns metadata for one exchange.

Required arguments

Use when

• You already know the venue and want its display metadata.
• You want to validate one exchange_id before issuing market requests.

markets_get_active

Returns active market IDs for one exchange.

Required arguments

Optional arguments

Important

This tool returns only market IDs, not full market objects.

Use when

• You want a compact discovery step before querying current state.
• You need valid market_id values but not the full market metadata yet.

markets_get_history

Returns market objects for one exchange, including descriptive and pricing fields.

Required arguments

Optional arguments

Returns

Each market object can include fields such as:

• market_id
• title
• description
• outcome_name
• price
• status
• exchange_id
• outcome_type
• source_specific_data

Use when

• You need titles, descriptions, status, or last-known pricing.
• You want richer context than markets_get_active provides.

Market activity tools

activity_get_current

Returns the latest trade and latest quote for one market.

Required arguments

Notes

• The tool description indicates either side of the payload may be null when data is unavailable.
• In practice, sparse markets may also return an explicit no-activity error.

Use when

• You need the freshest market activity signal.
• You are building quote cards, status monitors, or lightweight polling loops.

activity_get_latest

Returns a recent batch of trades and quotes for one market.

Required arguments

Use when

• You want more than just the single latest trade and quote.
• You are inspecting recent activity without going to flat-file history.

activity_get_quotes_history

Returns historical quote updates from flat files for one market.

Required arguments

Optional arguments

Important

• Results are filtered by data processing time, time_coinapi.
• Swagger describes the response as being in ascending time order.
• Use narrow windows for active markets to avoid large payloads.

Returns

Each row follows the quote tick schema and can include:

• entry_time
• recv_time
• ask
• bid
• ask_volume
• bid_volume

Use when

• You need historical bid/ask evolution rather than current state.
• You are replaying quote history for analytics or monitoring.

activity_get_trades_history

Returns historical trades from flat files for one market.

Required arguments

Optional arguments

Important

• Results are filtered by data processing time, time_coinapi.
• Prefer a small limit or a bounded time window when exploring.

Use when

• You need trade-level history rather than summary candles.
• You want to study execution flow or build time-and-sales style views.

OHLCV tools

ohlcv_list_periods

Returns the list of supported candle granularities.

Arguments

This tool does not require any arguments.

Typical periods

• 1SEC
• 1MIN
• 1HRS
• 1DAY
• 1MTH

Each row also includes metadata such as length_seconds, length_months, unit_count, unit_name, and display_name.

ohlcv_get_latest_market

Returns the latest OHLCV candles for one market in descending time order.

Required arguments

Optional arguments

Returns

Rows can include fields such as:

• time_period_start
• time_period_end
• time_open
• time_close
• price_open
• price_high
• price_low
• price_close
• volume_traded
• trades_count

ohlcv_get_history_market

Returns historical OHLCV candles for one market in ascending time order.

Required arguments

Optional arguments

Use when

• You need a bounded historical window for one market.
• You want deterministic ascending output for backfills or batch processing.

ohlcv_get_history_exchange

Returns historical OHLCV data for all markets on one exchange within a bounded time range.

Required arguments

Important

The time span between time_start and time_end cannot be more than 24 hours.

Use when

• You need venue-wide candles instead of a single market.
• You are building cross-sectional or exchange-level analytics for a short window.

Order book tools

orderbook_get_current

Returns the current order book snapshot for one market outcome.

Required arguments

Returns

The response can include:

• exchange_id
• market_id
• time_exchange
• time_coinapi
• asks
• bids

Each asks or bids row includes:

• price
• size

Use when

• You need the current liquidity ladder for one market outcome.
• You are computing spread, depth, or order book imbalance.

orderbook_get_history

Returns historical limit order updates from flat files for one market.

Required arguments

Optional arguments

Important

• This tool returns raw order updates, not reconstructed snapshots.
• Use small windows and explicit limits for active markets.

Use when

• You need order add, update, or delete events for replay.
• You are reconstructing book state in your own downstream system.

Practical examples

Example: list supported exchanges

{
  "tool": "exchanges_list",
  "arguments": {}
}

Example: active market IDs on Polymarket

{
  "tool": "markets_get_active",
  "arguments": {
    "exchange_id": "POLYMARKET",
    "limit": 100,
    "page": 1
  }
}

Example: rich market objects for one venue

{
  "tool": "markets_get_history",
  "arguments": {
    "exchange_id": "POLYMARKET",
    "limit": 25,
    "page": 1
  }
}

Example: latest daily candles for one market

{
  "tool": "ohlcv_get_latest_market",
  "arguments": {
    "exchange_id": "POLYMARKET",
    "market_id": "1-MEGATON-METEOR-STRIKE-IN-2026_NO",
    "period_id": "1DAY",
    "limit": 2
  }
}

Example: historical market candles for a fixed window

{
  "tool": "ohlcv_get_history_market",
  "arguments": {
    "exchange_id": "POLYMARKET",
    "market_id": "1-MEGATON-METEOR-STRIKE-IN-2026_NO",
    "period_id": "1DAY",
    "time_start": "2026-04-01T00:00:00Z",
    "time_end": "2026-04-07T00:00:00Z",
    "limit": 100
  }
}

Example: current order book snapshot

{
  "tool": "orderbook_get_current",
  "arguments": {
    "exchange_id": "POLYMARKET",
    "market_id": "1-MEGATON-METEOR-STRIKE-IN-2026_NO"
  }
}

Example: bounded trades history

{
  "tool": "activity_get_trades_history",
  "arguments": {
    "exchange_id": "POLYMARKET",
    "market_id": "1-MEGATON-METEOR-STRIKE-IN-2026_NO",
    "time_start": "2026-04-07T12:00:00Z",
    "time_end": "2026-04-07T13:00:00Z",
    "limit": 100
  }
}

Example: bounded quotes history

{
  "tool": "activity_get_quotes_history",
  "arguments": {
    "exchange_id": "POLYMARKET",
    "market_id": "1-MEGATON-METEOR-STRIKE-IN-2026_NO",
    "time_start": "2026-04-07T13:35:00Z",
    "time_end": "2026-04-07T13:40:00Z",
    "limit": 5
  }
}

Example response shape:

[
  {
    "entry_time": "2026-04-07T13:35:12.0000000Z",
    "recv_time": "2026-04-07T13:35:12.1230000Z",
    "ask": 0.949,
    "bid": 0.928,
    "ask_volume": 112.39,
    "bid_volume": 81.55
  }
]

Usage guidance

• Start with exchanges_list and markets_get_active when you need quick discovery with small payloads.
• Use markets_get_history when you need human-readable titles, descriptions, and price context.
• Call ohlcv_list_periods before issuing candle requests from a new client or workflow.
• Use activity_get_current and orderbook_get_current for low-latency monitoring.
• Prefer bounded time_start / time_end windows over full-day historical pulls when exploring flat-file tools.
• Remember that orderbook_get_history returns raw updates, so consumers usually need their own reconstruction logic.