KnownAt API

Liveness probe

Reports whether the API Worker is running. This endpoint performs no dependency checks and no authentication, so it stays answerable during a database outage and is safe for a high-frequency external uptime monitor. Use /health/ready to check dependencies and /v1/status to check data freshness.

Response · 200

The Worker is running.

status"ok" | "degraded"

Whether the Worker itself is serving requests.

servicestring

environmentstring

releasestring

Release identifier of the running deployment, for correlating errors with a deploy.

git_shastring

generated_atstring (date-time)

UTC timestamp in ISO 8601 format with millisecond precision.

Errors

Readiness probe

Probes every dependency required to serve API traffic and reports each one individually with its latency. Returns 503 when any dependency is unhealthy, so a load balancer or deployment gate can act on it. Slower than /health: do not poll it at high frequency.

Response · 200

Every dependency is healthy.

status"ok" | "degraded"

Whether the Worker itself is serving requests.

servicestring

environmentstring

releasestring

Release identifier of the running deployment, for correlating errors with a deploy.

git_shastring

generated_atstring (date-time)

UTC timestamp in ISO 8601 format with millisecond precision.

dependencies[]object[]

Result of probing each dependency required to serve requests.

dependencies[].namestring

dependencies[].healthyboolean

dependencies[].latency_msinteger

dependencies[].errorstring | null

Errors

Ingestion pipeline health

Flat view of every collector heartbeat, including source, receive and write lag per feed. Intended for the KnownAt watchdog and operational dashboards rather than for client applications, which should use /v1/status. Requires the account:read scope.

Response · 200

Current state of every ingestion feed.

status"operational" | "delayed" | "degraded" | "offline" | "maintenance"

`operational` — the feed is within its documented lag budget. `delayed` — data is still arriving but later than the budget allows. `degraded` — lag is severe enough to affect most use cases. `offline` — no recent data has been persisted. `maintenance` — a planned interruption, announced in advance.

generated_atstring (date-time)

UTC timestamp in ISO 8601 format with millisecond precision.

feeds[]object[]

Flat list of every collector heartbeat, for internal dashboards and the watchdog.

feeds[].status"operational" | "delayed" | "degraded" | "offline" | "maintenance"

`operational` — the feed is within its documented lag budget. `delayed` — data is still arriving but later than the budget allows. `degraded` — lag is severe enough to affect most use cases. `offline` — no recent data has been persisted. `maintenance` — a planned interruption, announced in advance.

feeds[].freshness_basis"persisted_observations"

For Predict.fun, timestamps describe queryable observations. Receive time is confirmed through storage, not an independent live recorder heartbeat. Catalog has no source event clock. Rates count projected rows across five complete receive-time minutes.

feeds[].source_lag_msinteger | null

Time since the newest event timestamp reported by the source. Grows when the source itself stops producing data.

feeds[].receive_lag_msinteger | null

Time since KnownAt last received an event. Grows when the collector loses its connection to the source.

feeds[].write_lag_msinteger | null

Time since KnownAt last persisted an event. Fresh receive lag with stale write lag means the pipeline, not the source, is failing.

feeds[].last_source_timestampstring (date-time) | null

Event time of the newest record, as reported by the source.

feeds[].last_received_atstring (date-time) | null

When KnownAt received that record.

feeds[].last_written_atstring (date-time) | null

When KnownAt persisted that record to queryable storage.

feeds[].events_per_secondnumber

Recent throughput of the feed.

feeds[].feedstring

Feed identifier, `dataset.stream`.

feeds[].datasetstring

feeds[].streamstring

Errors

Data freshness and feed status

Current status and lag of every live KnownAt feed, grouped by dataset and stream. Public and unauthenticated so it can be polled by a client before it decides to trust live data, or wired into a customer-side monitor. status at the top level is the worst status across all feeds. Poll at most once every 10 seconds.

Response · 200

Status of every feed.

status"operational" | "delayed" | "degraded" | "offline" | "maintenance"

Worst status across every feed. Use it as a single health signal.

generated_atstring (date-time)

UTC timestamp in ISO 8601 format with millisecond precision.

datasetsobject

Per-dataset, per-stream freshness, keyed by dataset slug and then by stream name.

datasets.{key}.{key}.status"operational" | "delayed" | "degraded" | "offline" | "maintenance"

`operational` — the feed is within its documented lag budget. `delayed` — data is still arriving but later than the budget allows. `degraded` — lag is severe enough to affect most use cases. `offline` — no recent data has been persisted. `maintenance` — a planned interruption, announced in advance.

datasets.{key}.{key}.freshness_basis"persisted_observations"

For Predict.fun, timestamps describe queryable observations. Receive time is confirmed through storage, not an independent live recorder heartbeat. Catalog has no source event clock. Rates count projected rows across five complete receive-time minutes.

datasets.{key}.{key}.source_lag_msinteger | null

Time since the newest event timestamp reported by the source. Grows when the source itself stops producing data.

datasets.{key}.{key}.receive_lag_msinteger | null

Time since KnownAt last received an event. Grows when the collector loses its connection to the source.

datasets.{key}.{key}.write_lag_msinteger | null

Time since KnownAt last persisted an event. Fresh receive lag with stale write lag means the pipeline, not the source, is failing.

datasets.{key}.{key}.last_source_timestampstring (date-time) | null

Event time of the newest record, as reported by the source.

datasets.{key}.{key}.last_received_atstring (date-time) | null

When KnownAt received that record.

datasets.{key}.{key}.last_written_atstring (date-time) | null

When KnownAt persisted that record to queryable storage.

datasets.{key}.{key}.events_per_secondnumber

Recent throughput of the feed.

reconciliationobject

Independent evidence coverage. Computed over the latest 15 complete non-empty minute windows after finality and lateness grace.

reconciliation.polymarket_wallet_tradesobject

reconciliation.polymarket_wallet_trades.status"operational" | "delayed" | "degraded" | "offline" | "maintenance"

`operational` — the feed is within its documented lag budget. `delayed` — data is still arriving but later than the budget allows. `degraded` — lag is severe enough to affect most use cases. `offline` — no recent data has been persisted. `maintenance` — a planned interruption, announced in advance.

reconciliation.polymarket_wallet_trades.window_startstring (date-time) | null

UTC timestamp in ISO 8601 format with millisecond precision.

reconciliation.polymarket_wallet_trades.window_endstring (date-time) | null

UTC timestamp in ISO 8601 format with millisecond precision.

reconciliation.polymarket_wallet_trades.computed_atstring (date-time) | null

UTC timestamp in ISO 8601 format with millisecond precision.

reconciliation.polymarket_wallet_trades.data_api_rowsinteger

reconciliation.polymarket_wallet_trades.data_api_transactionsinteger

reconciliation.polymarket_wallet_trades.matched_rowsinteger

reconciliation.polymarket_wallet_trades.ambiguous_rowsinteger

reconciliation.polymarket_wallet_trades.transaction_only_rowsinteger

reconciliation.polymarket_wallet_trades.missing_onchain_rowsinteger

reconciliation.polymarket_wallet_trades.invalid_hash_rowsinteger

reconciliation.polymarket_wallet_trades.matched_transactionsinteger

reconciliation.polymarket_wallet_trades.exact_row_coveragenumber

Share of Data API rows matched to exactly one economically identical canonical OrderFilled.

reconciliation.polymarket_wallet_trades.transaction_coveragenumber

Share of unique Data API transaction hashes present in canonical Polygon OrderFilled events.

reconciliation.polymarket_wallet_trades.max_finalized_blockinteger

Errors

Verify an API key

Returns the identity, scopes and plan associated with the presented API key. Use it as a connectivity and credentials check when wiring up a client: a 200 means the key is valid, unexpired and unrevoked, and the scopes array tells you exactly what it may call. The key itself is never echoed back.

Response · 200

The key is valid.

authenticatedtrue

keyobject

Everything the API knows about the presented key. The key itself is never echoed.

key.idstring (uuid)

Identifier of the presented key, safe to log.

key.namestring

Label given to the key in the dashboard.

key.environment"test" | "live"

Whether this is a test or live key, derived from its prefix.

key.scopes[]string[]

Permissions granted to this key.

key.planstring

key.expires_atstring (date-time) | null

Expiry, or null for a non-expiring key.

Errors

Current period usage

Exact subscription usage and live database limits for the configured quota period. Successful data queries (API and Explorer) consume requests, returned rows and JSON bandwidth. Account/catalog reads are free. Pending requests reserve allowance for up to two minutes. Includes daily usage and dataset/endpoint activity. The response example is illustrative; actual limits are read from the subscription configuration.

Response · 200

Usage for the current period.

planstring

plan_namestring

can_queryboolean

periodobject

Quota period from subscription plan configuration. Billing periods follow Paddle; calendar-month plans reset at UTC month boundaries.

period.startstring (date-time)

UTC timestamp in ISO 8601 format with millisecond precision.

period.endstring (date-time)

UTC timestamp in ISO 8601 format with millisecond precision.

requestsobject

Successful product requests plus in-flight reservations. Account and catalog reads are excluded.

requests.usednumber

requests.limitnumber

rowsobject

Product rows returned, not database rows scanned.

rows.usednumber

rows.limitnumber

bandwidth_gbobject

Uncompressed UTF-8 JSON response size in GiB (2^30 bytes).

bandwidth_gb.usednumber

bandwidth_gb.limitnumber

limitsobject

limits.requests_per_minutenumber

limits.expensive_requests_per_minutenumber

limits.max_page_sizenumber

limits.max_api_keysnumber

limits.quota_resetstring

daily[]object[]

daily[].daystring

daily[].requestsnumber

daily[].rowsnumber

daily[].bytesnumber

datasets[]object[]

datasets[].datasetstring

datasets[].endpointstring

datasets[].requestsnumber

datasets[].rowsnumber

datasets[].last_used_atstring (date-time)

UTC timestamp in ISO 8601 format with millisecond precision.

Errors

Read precomputed aggregate market correlations

Pearson correlation of aligned aggregate matched-venue log returns. No filling gaps. At least 60% of the full window plus a minimum number of pairs is required. Results refresh in the background every minute. No causality or identity claim.

Parameters

horizonstring · default 1hoptional

Return window the correlation is computed over.

leftstringoptional

Canonical ID of the first market of a pair. Requires right; returns that pair's recorded history.

rightstringoptional

Canonical ID of the second market of a pair. Requires left.

Response · 200

Graph, quality gates and recorded correlation history

Errors

Read aggregate reference candles and current snapshot

Recorded reference candles for the active publisher generation: the last twelve hours of one-minute candles, or up to 720 candles of a longer interval rolled up from them, the forming candle included through the last completed minute. Flows sum; CVD, open interest and depth are end-of-candle states. Gaps remain missing. Requires product access.

Parameters

marketstringrequired

Canonical ID of the aggregated market.

intervalstring · default 1moptional

Candle interval; buckets are aligned to UTC.

beforeintegeroptional

Older page: up to 720 recorded candles before this candle start, in epoch milliseconds, aligned to the interval. Returns more instead of through and snapshot.

Response · 200

Recorded reference candles

generationstring

marketstring

intervalstring

throughnumber

Last recorded minute included, in epoch milliseconds.

moreboolean

Older page only: whether earlier candles remain.

candles[]object[]

snapshotobject

Errors

Read captured order book liquidity by distance band

Captured resting liquidity for an aggregated market: the last 720 candles of the interval, one frame per candle (the last captured minute in it), summed in USD across sources into distance bands from the reference price: 0.02% bands within 0.1%, then wider bands out to 10%. Resting orders, not executions; coverage varies by exchange. Requires product access.

Parameters

marketstringrequired

Canonical ID of the aggregated market.

intervalstring · default 1moptional

Candle interval; buckets are aligned to UTC.

beforeintegeroptional

Older page: up to 720 frames before this candle start, in epoch milliseconds, aligned to the interval. Adds more.

Response · 200

Liquidity frames

generationstring

marketstring

intervalstring

bands[]unknown[][]

minutes[]number[]

references[]number[]

cells[]unknown[][]

moreboolean

Older page only: whether earlier frames remain.

Errors

Open a five-minute authenticated reference-price stream

One-use WebSocket ticket for admitted aggregate markets. Omit market for quotes-only directory updates; specify market for its candles and book. Requires product access; quotes are indicative, not executable. Maximum four connections per workspace. Reauthenticate after expiry.

Parameters

marketstringoptional

Canonical ID of the aggregated market to stream candles and book for. Omit for quotes-only directory updates.

Response · 200

Single-use stream ticket

ticketstring

expires_atnumber

urlstring

Errors

Discover captured instruments

Shared directory for the workspace and API clients. Currently covers Hyperliquid metadata captured by KnownAt, including inactive instruments. Returns the latest metadata per source instrument at a fixed ingestion cutoff, without a 24-hour discovery window. Presence does not certify active trading or complete history. Source identities are not canonical assets; classification and cross-venue mapping are separate. Search matches source symbol and available source name/coin fields. Pagination is signed and bound to search, cutoff and the market-cap snapshot. Capitalization is a periodically refreshed local snapshot for explicitly mapped native coins; unmapped instruments sort last. Metadata is source-native, not a stable normalized asset schema.

Parameters

directionstringoptional

Sort direction. Missing values and inactive contracts remain last.

sortstringoptional

Global sort before pagination, inactive contracts last. Defaults to volume for TradFi categories, market cap for all/crypto. Volume ranks recent captured 24h quote-volume; missing/stale values last within each status group.

overviewstringoptional

Include last captured market context, including historical values for inactive contracts. observed_at is the snapshot time; these are not necessarily current values. Missing coverage is null.

source_idstringoptional

Exact source instrument identifier, case-sensitive.

marketstring · default alloptional

Categories use explicit source-contract mappings. TradFi includes non-crypto HIP-3 contracts, including unclassified new listings. Unclassified filters HIP-3 contracts awaiting classification. Crypto includes native and mapped HIP-3 crypto markets.

qstring · default optional

Case-insensitive source symbol or available source name.

limitinteger · default 50optional

Maximum instruments per page; up to 100.

cursorstringoptional

Signed cursor bound to the search and ingestion cutoff.

Response · 200

Instrument metadata, ordered globally by the requested metric with inactive contracts last and source instrument ID as a tie-breaker. Snapshot values may be historical; inspect observed_at.

data[]object[]

data[].snapshotobject | null

data[].snapshot.pricenumber | null

data[].snapshot.change_24hnumber | null

data[].snapshot.volume_24hnumber | null

data[].snapshot.funding_ratenumber | null

data[].snapshot.open_interestnumber | null

data[].snapshot.observed_atstring

data[].capitalizationobject | null

data[].capitalization.usdnumber

data[].capitalization.updated_atstring

data[].idstring

data[].source"hyperliquid"

data[].source_instrument_idstring

data[].metadata_observed_atstring

data[].display_namestring

data[].symbolstring

data[].kind"spot" | "perpetual" | "unknown"

data[].category"crypto" | "stocks" | "metals" | "energy" | "agriculture" | "indices" | "forex" | "rates" | "funds" | "pre-ipo" | "unclassified"

data[].venuestring

data[].quotestring | null

data[].status"active" | "inactive" | "unknown"

data[].logostring | null

data[].metadataobject

metaobject

meta.direction"asc" | "desc"

meta.sort"name" | "price" | "change" | "market_cap" | "volume" | "funding" | "open_interest"

meta.inactive_lastboolean

meta.recordsnumber

meta.has_moreboolean

meta.next_cursorstring | null

meta.as_ofstring

meta.generated_atstring

meta.sources[]string[]

meta.market_capobject

meta.market_cap.sourcestring

meta.market_cap.captured_atstring

meta.market_cap.orderstring

meta.market_cap.missingstring

Errors

Browse canonical assets and cross-venue contracts

Captured Hyperliquid, Bybit and Deribit identities. mode=assets lists verified asset groups with a verified Hyperliquid perpetual where one exists and, for assets with an aggregate reference, its trailing-day summary (last close, change since the window start, captured perpetual volume; the window starts later when the reference generation is younger); mode=instruments requires asset and separates verified mappings from candidates; mode=review prioritizes unresolved contracts with recent trades/books. Spot, futures, perpetuals and options retain distinct identities. Mapping never grants aggregation eligibility. Coverage endpoints do not prove continuous history. Cursors expire when the published catalog revision changes.

Parameters

modestring · default assetsoptional

List verified assets, contracts of one asset, or the mapping review queue.

assetstringoptional

Canonical asset ID, required for instruments; for example metals:gold.

qstring · default optional

Case-insensitive asset name, symbol or source contract search.

categorystringoptional

Filter by asset class; tradfi selects every class except crypto.

contractstringoptional

Exact catalog contract ID, for example hyperliquid:BTC. In assets mode, returns the asset that contract is verified to reference.

sourcestringoptional

Exchange filter for contract and review modes.

kindstringoptional

Contract type filter for contract and review modes.

statusstringoptional

Last source-reported status; expiry and data recency are separate.

mappingstring · default verifiedoptional

For instruments, verified relationships or proposed mappings awaiting review.

limitinteger · default 40optional

Maximum records per page.

cursorstringoptional

Signed cursor bound to filters and catalog revision. Refresh without a cursor after a catalog update.

Response · 200

Persisted canonical catalog page

data[]object[]

assetobject | null

metaobject

meta.revisionstring

meta.updated_atstring

meta.generated_atstring

meta.recordsnumber

meta.total_unfilterednumber | null

meta.has_moreboolean

meta.next_cursorstring | null

meta.aggregation_eligiblefalse

meta.coverage_continuitystring

Errors

Page captured chart candles backwards

At most 500 candle slots per request, aggregated from deduplicated captured trades at display precision. Missing slots are omitted. Pass next_before and the original as_of for older history. Values use display precision and are intended for product charts; internal exact-decimal source records are not exposed.

Parameters

market_idstringrequired

Exact Hyperliquid source contract identifier.

intervalstring · default 1moptional

Candle duration; up to 500 aligned time slots per page.

beforestringoptional

Exclusive UTC candle boundary returned as next_before.

as_ofstringoptional

Fixed ingestion cutoff from the first page. Defaults to the current completed minute.

Response · 200

Bounded candle page.

rows[]object[]

rows[].observed_atstring

rows[].opennumber

rows[].highnumber

rows[].lownumber

rows[].closenumber

rows[].volumenumber

rows[].vwapnumber | null

sincestring

untilstring

as_ofstring

next_beforestring | null

marketstring

interval"1m" | "5m" | "15m" | "1h" | "4h" | "1d"

interval_secondsnumber

Errors

Discover captured prediction markets and recurring series

Latest captured catalog for Polymarket and predict.fun. Groups recurring crypto series and source-linked candidates for navigation; this does not certify equivalent settlement rules. Contracts retain venue identities. Category is based on source tags; unknown categories remain Other. Status is last captured status, not a live trading guarantee. Live keyset pagination, not a historical snapshot. Group selects individual contracts. No price or model data is fabricated.

Parameters

qstring · default optional

Search captured title or source slug.

categorystring · default alloptional

Category derived from source tags.

venuestring · default alloptional

Filter by captured source venue.

statusstring · default openoptional

Last captured status. Open excludes elapsed windows with known schedules.

limitinteger · default 30optional

Page size, from 1 to 100.

cursorstringoptional

Signed keyset cursor bound to filters.

groupstringoptional

Exact navigation group ID to list its individual contracts.

Response · 200

Paged groups or venue contracts, ordered by stable identity.

data[]object[]

metaobject

meta.modestring

meta.has_moreboolean

meta.next_cursorstring | null

meta.generated_atstring

meta.cache"hit" | "miss" | "stale" | "coalesced"

Shared catalog cache status. generated_at is the snapshot time, not the request time.

Errors

Captured prices and book for a prediction contract

One venue contract, up to 24 hours and 601 sampled points per outcome. Books remain venue-specific. Midpoint is calculated only from valid two-sided uncrossed quotes; missing quotes remain null. Polymarket price snapshots are an explicitly labelled fallback when no books are captured. Predict.fun outcome indexSet=2 is derived by swapping and complementing the Yes book with source precision. Quote timestamps are preserved; this is captured data, not a live execution feed.

Parameters

venuestringrequired

Source venue.

market_idstringrequired

Exact venue catalog market ID, not a group ID.

sincestringrequired

Inclusive UTC window start.

untilstringrequired

Inclusive UTC end, at most 24 hours after since.

Response · 200

Bounded captured outcome prices and the last book within the window.

outcomes[]object[]

metaobject

Errors

Up/Down probability ladder across horizons for one asset

For each horizon (5m, 15m, 1h, 4h, 1d) the market whose window is open and the next one: probability, change over 5 and 15 minutes, a z-score against the same asset, horizon and point in the window over 30 days, and 15-minute traded volume with a thin flag. Polymarket's recurring Up or Down markets, priced from their trades (Up price; a Down trade at q counts as 1 - q). Trades arrive 2-5 minutes after they happen.

Parameters

assetstringrequired

Asset ticker: BTC, ETH, SOL, XRP, DOGE, BNB, HYPE or ZEC for 5m-1d crypto windows; stock and index tickers (for example NVDA, SPX) list daily windows.

Response · 200

One row per horizon.

assetstring

data[]object[]

data[].horizonstring

data[].horizon_secondsnumber

data[].liveobject | null

data[].live.condition_idstring

data[].live.market_idstring

data[].live.slugstring

data[].live.starts_atstring

data[].live.ends_atstring

data[].live.elapsednumber | null

Share of the window already elapsed, 0 before it opens.

data[].live.probabilitynumber | null

Up price of the last traded minute.

data[].live.price_atstring | null

data[].live.change_5mnumber | null

data[].live.change_15mnumber | null

data[].live.phasenumber

-1 before the window opens, then 0-9 for each tenth of it.

data[].live.znumber | null

Standard deviations from the mean price of the same asset, horizon and phase over 30 days.

data[].live.typicalobject | null

data[].live.typical.meannumber | null

data[].live.typical.sdnumber | null

data[].live.typical.markets_minutesnumber

data[].live.usdc_15mnumber | null

data[].live.trades_15mnumber

data[].live.thinboolean

Less than meta.thin_usdc traded in the last 15 minutes: the price is unreliable.

data[].live.staleboolean

No trade within meta.stale_seconds.

data[].nextobject | null

data[].next.condition_idstring

data[].next.market_idstring

data[].next.slugstring

data[].next.starts_atstring

data[].next.ends_atstring

data[].next.elapsednumber | null

Share of the window already elapsed, 0 before it opens.

data[].next.probabilitynumber | null

Up price of the last traded minute.

data[].next.price_atstring | null

data[].next.change_5mnumber | null

data[].next.change_15mnumber | null

data[].next.phasenumber

-1 before the window opens, then 0-9 for each tenth of it.

data[].next.znumber | null

Standard deviations from the mean price of the same asset, horizon and phase over 30 days.

data[].next.typicalobject | null

data[].next.typical.meannumber | null

data[].next.typical.sdnumber | null

data[].next.typical.markets_minutesnumber

data[].next.usdc_15mnumber | null

data[].next.trades_15mnumber

data[].next.thinboolean

Less than meta.thin_usdc traded in the last 15 minutes: the price is unreliable.

data[].next.staleboolean

No trade within meta.stale_seconds.

metaobject

meta.generated_atstring

meta.history_daysnumber

meta.thin_usdcnumber

meta.stale_secondsnumber

meta.notestring

Errors

Up probability per horizon over time

Rows for a time x horizon heatmap: per bucket, the Up price at the last traded minute of the market whose window was open. Polymarket's recurring Up or Down markets, priced from their trades (Up price; a Down trade at q counts as 1 - q). Trades arrive 2-5 minutes after they happen.

Parameters

assetstringrequired

Asset ticker: BTC, ETH, SOL, XRP, DOGE, BNB, HYPE or ZEC for 5m-1d crypto windows; stock and index tickers (for example NVDA, SPX) list daily windows.

sincestringoptional

Inclusive start. Defaults to 24 hours before until.

untilstringoptional

Exclusive end. Defaults to now.

intervalinteger · default 300optional

Bucket width in seconds; at most 2,000 buckets per request.

Response · 200

Points per horizon, oldest first.

assetstring

data[]object[]

data[].horizonstring

data[].horizon_secondsnumber

data[].points[]object[]

data[].points[].timenumber

data[].points[].probabilitynumber | null

data[].points[].usdcnumber | null

metaobject

meta.sincestring

meta.untilstring

meta.interval_secondsnumber

meta.notestring

Errors

How often Up/Down forecasts came true

Resolved markets binned by forecast (5 points wide) with the share that resolved Up, Brier skill against a coin flip and the calibration slope, per horizon. Polymarket's recurring Up or Down markets, priced from their trades (Up price; a Down trade at q counts as 1 - q). Trades arrive 2-5 minutes after they happen.

Parameters

assetstringoptional

Asset ticker; every asset when omitted.

atstring · default openoptional

open: the price in the two minutes before the window opens; half: in the window's fifth tenth.

sincestringoptional

Windows ending at or after this instant. Defaults to 30 days ago.

Response · 200

Calibration per horizon.

assetstring | null

at"open" | "half"

data[]object[]

data[].horizonstring

data[].horizon_secondsnumber

data[].marketsnumber

data[].up_ratenumber | null

data[].briernumber | null

data[].brier_coinnumber

data[].skillnumber | null

1 - Brier / 0.25: share of the coin-flip error removed.

data[].slopenumber | null

Outcome on forecast distance from 50%: 1 calibrated, above 1 under-confident.

data[].bins[]object[]

data[].bins[].fromnumber

data[].bins[].tonumber

data[].bins[].marketsnumber

data[].bins[].forecastnumber | null

data[].bins[].realizednumber | null

metaobject

meta.sincestring

meta.untilstring

meta.notestring

Errors

Read published Whale Alert v2 revisions

Immutable emission, correction and retraction stream ordered by publication time (newest first). Uses stored whale-alert.v2 decisions, never recalculates historical alerts. as_of freezes publication visibility; next_cursor preserves that snapshot and filters. Apply revisions by alert_key and revision, not transaction time. A minimum-notional filter retains subsequent retractions of matching alerts. Returns preparing until the explicit public release. Requires polymarket:read or predict_fun:read according to venue. Polymarket and Predict.fun via the venue parameter, with independent calibration and release. Predict.fun uses whale-alert.predict-fun.v2 with identical numerical rules and BNB settlement evidence; unusual position building does not establish smart-money identity or profitability.

Parameters

venuestringoptional

Venue. Defaults to polymarket; predict_fun has its own calibration, immutable source manifest and release gate.

as_ofstringoptional

Publication cutoff. Omit for current visibility; pagination freezes this instant.

limitinteger · default 20optional

Maximum published revisions per page, between 1 and 100.

cursorstringoptional

Exact next_cursor string from the previous page, using unchanged filters and cutoff.

walletstringoptional

Exact lowercase wallet address.

marketstringoptional

Exact lowercase condition ID on the selected venue.

min_net_usdnumbernulloptional

Minimum net buying of a published alert. Later corrections and withdrawals of matching alerts remain visible.

Response · 200

Stored revisions or explicit preparing state.

data[]object[]

data[].alert_keystring

data[].revisioninteger

data[].action"emitted" | "corrected" | "retracted"

data[].published_atstring

data[].input_cutoffstring

data[].calibration_fingerprintstring

data[].payloadobject

data[].payload.candidateobject

data[].payload.candidate.venuestring

data[].payload.candidate.walletstring

data[].payload.candidate.marketstring

data[].payload.candidate.tokenstring

data[].payload.candidate.startstring

data[].payload.candidate.endstring

data[].payload.candidate.buyUsdstring

data[].payload.candidate.sellUsdstring

data[].payload.candidate.netUsdstring

data[].payload.candidate.netSharesstring

data[].payload.candidate.availableAtstring

data[].payload.candidate.executionIds[]string[]

data[].payload.decisionstring

data[].payload.reasons[]string[]

data[].payload.smart_money"unverified"

data[].payload.wallet_profilestring

data[].payload.concurrent_outcome_tokens[]string[]

data[].payload.decision_fingerprintstring

metaobject

meta.status"preparing" | "live"

meta.signal_version"whale-alert.v2" | "whale-alert.predict-fun.v2"

meta.as_ofstring

meta.recordsinteger

meta.has_moreboolean

meta.next_cursorstring | null

meta.first_windowstring | null

meta.can_filterboolean

meta.can_replayboolean

Errors

Read published Accumulation v2 revisions

Immutable emission, correction and retraction stream ordered by publication time (newest first). Uses stored accumulation.v2 decisions, never recalculates historical alerts. as_of freezes publication visibility; next_cursor preserves that snapshot and filters. Apply revisions by alert_key and revision, not transaction time. A minimum-notional filter retains subsequent retractions of matching alerts. Returns preparing until the explicit public release. Polymarket and Predict.fun use the same numerical rules over 24h, 72h and 168h. Select venue=predict_fun for BNB evidence and accumulation.predict-fun.v2 history. Requires predict_fun:read or polymarket:read according to venue; sustained buying does not establish smart-money identity or profitability.

Parameters

venuestringoptional

Venue. Defaults to polymarket. Predict.fun has independent BNB evidence, publication history and release.

as_ofstringoptional

Publication cutoff. Omit for current visibility; pagination freezes this instant.

limitinteger · default 20optional

Maximum published revisions per page, between 1 and 100.

cursorstringoptional

Exact next_cursor string from the previous page, using unchanged filters and cutoff.

walletstringoptional

Exact lowercase wallet address.

marketstringoptional

Exact lowercase condition ID on the selected venue.

min_net_usdnumbernulloptional

Minimum net buying of a published alert. Later corrections and withdrawals of matching alerts remain visible.

Response · 200

Stored revisions or explicit preparing state.

data[]object[]

data[].alert_keystring

data[].revisioninteger

data[].action"emitted" | "corrected" | "retracted"

data[].published_atstring

data[].input_cutoffstring

data[].calibration_fingerprintstring

data[].payloadobject

data[].payload.candidateobject

data[].payload.candidate.venuestring

data[].payload.candidate.walletstring

data[].payload.candidate.marketstring

data[].payload.candidate.tokenstring

data[].payload.candidate.startstring

data[].payload.candidate.endstring

data[].payload.candidate.asOfstring

data[].payload.candidate.netUsdstring

data[].payload.candidate.netSharesstring

data[].payload.candidate.boughtUsdstring

data[].payload.candidate.soldUsdstring

data[].payload.candidate.marketBoughtUsdstring

data[].payload.candidate.marketBuyShareBpsnumber

data[].payload.candidate.buyCountnumber

data[].payload.candidate.buyHoursnumber

data[].payload.candidate.buySpanMsnumber

data[].payload.candidate.executions[]object[]

data[].payload.candidate.executions[].idstring

data[].payload.candidate.executions[].venuestring

data[].payload.candidate.executions[].walletstring

data[].payload.candidate.executions[].marketstring

data[].payload.candidate.executions[].tokenstring

data[].payload.candidate.executions[].side"BUY" | "SELL"

data[].payload.candidate.executions[].eventAtstring

data[].payload.candidate.executions[].availableAtstring

data[].payload.candidate.executions[].usdstring

data[].payload.candidate.executions[].sharesstring

data[].payload.decisionstring

data[].payload.reasons[]string[]

data[].payload.decision_fingerprintstring

metaobject

meta.status"preparing" | "live"

meta.signal_version"accumulation.v2" | "accumulation.predict-fun.v2"

meta.as_ofstring

meta.recordsinteger

meta.has_moreboolean

meta.next_cursorstring | null

meta.first_windowstring | null

meta.can_filterboolean

meta.can_replayboolean

Errors

Read a verified Top Traders ranking snapshot

Polymarket and Predict.fun have independent evidence, versioned snapshots and public release gates. Select venue=predict_fun with predict_fun:read; defaults to polymarket with polymarket:read. Predict.fun uses top-traders.predict-fun.v2 and stays preparing until wallet costs, fee treatment and settlement are reconciled. Seven-day realized execution P&L after charged execution fees. Rankings are separate for each collateral token; amounts are nominal collateral units, not an FX-adjusted USD total. Rewards, fee rebates and unrealized price changes are excluded. ROI is return on realized cost, not capital at risk. At least five traded markets must be fully closed and realized cost must be at least 100 collateral units. Unknown cost or incomplete portfolio evidence excludes the portfolio, with coverage counts reported. as_of and next_cursor freeze the published snapshot and its revision. Corrections never rewrite earlier snapshots. Returns preparing until a validated public release. No sampled-position audit is eligible for this feed.

Parameters

venuestringoptional

Venue; defaults to polymarket. Predict.fun remains preparing until complete wallet cost and settlement evidence is verified.

as_ofstringoptional

Publication cutoff; omit for latest visibility. Historical replay requires an eligible subscription.

limitinteger · default 20optional

Maximum portfolios per page. Anonymous web access is limited to 20.

cursorstringoptional

Exact next_cursor from the previous page; preserves snapshot, revision, filters and publication cutoff.

walletstringoptional

Exact lowercase wallet address. Sign-in required.

collateralstringoptional

Exact lowercase collateral contract. Rankings and nominal amounts are separate per collateral.

Response · 200

Immutable ranking snapshot or preparing state

data[]object[]

data[].walletstring

data[].collateralstring

data[].rankinteger

data[].realized_pnlstring

data[].realized_coststring

data[].closed_marketsinteger

data[].return_on_realized_cost_bpsnumber

data[].open_positions[]object[]

data[].open_positions[].tokenstring

data[].open_positions[].sharesstring

data[].open_positions[].coststring | null

data[].proof_hashstring

metaobject

meta.status"preparing" | "live"

meta.signal_version"top-traders.v2" | "top-traders.predict-fun.v2"

meta.as_ofstring

meta.recordsinteger

meta.next_cursorstring | null

meta.has_moreboolean

meta.can_filterboolean

meta.can_replayboolean

meta.snapshot_keystring | null

meta.revisioninteger | null

meta.window_startstring | null

meta.window_endstring | null

meta.observed_portfoliosinteger

meta.eligible_portfoliosinteger

meta.excluded_portfoliosinteger

Errors

Compare conditional carry returns and token-matched buying context

Immutable yield.v2 input snapshots, with original yield.v1 replay, for Polymarket and Predict.fun. Calculates purchases against recorded ask depth for an entry budget. Payout assumes the selected outcome wins and pays one collateral unit per share; price is not a calibrated probability. Partial depth reports unspent capital. Polymarket entry-fee estimates use the stored feeSchedule with exponent 1 and the pinned SDK cash-fee curve; zero-fee flags are supported. Predict v2 uses each stored market feeRateBps and deducts the base taker fee from purchased shares: floor(bps*min(p,1-p)*shares/p/10000), using D18 price and rounding per observed level. Account-specific discounts and refunds are excluded. purchased_shares and payout_before_entry_fee_if_win are before the share fee; shares and payout_if_win are after it. estimated_entry_fee is the entry-price collateral equivalent of the fee, not an extra cash debit for Predict. Exact per-fill rounding can differ for aggregated depth. Unverified fees remain null. The budget includes known entry fees; gas, funding, withdrawal, redemption and optional builder costs are excluded. Predict scheduled ends come from captured category endsAt with exact market/condition membership and a 15-minute freshness limit; missing dates remain unknown. Scheduled end is not promised settlement. Annualization is simple, not compounded; Predict collateral yield/rewards are excluded. Published Whale Alert and Accumulation context matches venue, condition and token at the snapshot cutoff; purchases are not proof of current holdings. Polymarket Top Traders context uses the last published holdings from top 100 portfolios per collateral, at most 24 hours old, and never claims a current balance. Unreleased rankings remain preparing. Predict holdings stay unavailable until their CTF contract identity is supported. Snapshots refresh every five minutes, cover a bounded candidate sample, and retain publication history for replay. An omitted token in a later snapshot means it is no longer in that candidate sample, not that a prior trade was reversed. Scopes are polymarket:read or predict_fun:read by venue.

Parameters

venuestring · default polymarketoptional

Venue and collateral denomination.

categorystring · default alloptional

Filter by source or inferred market category; all disables category filtering.

budgetstring · default 100optional

Entry budget in venue collateral units, including verified entry fees; unverified fees and external costs are excluded. Default 100; custom budget requires sign-in.

min_return_pctnumbernull · default 0optional

Minimum conditional gross return percentage. Not an estimated win probability.

max_daysintegeroptional

Maximum days until scheduled end; unknown/overdue dates are excluded when set.

settlement_delay_daysintegernull · default 0optional

Additional assumed settlement delay for simple annualization.

contextstring · default anyoptional

Require available, token-matched published buying context. Unreleased indicators never count as zero activity.

sortstring · default returnoptional

Sort by conditional gross return, net return (unknown last), recorded capacity or scheduled end.

as_ofstringoptional

Publication cutoff for point-in-time replay; subscription required for historical cutoffs.

limitinteger · default 20optional

Page size; public access is capped at 20.

cursorstringoptional

Cursor binds venue, query, snapshot and cutoff. Use unchanged filters on subsequent pages.

Response · 200

Calculated carry page from a fixed stored input snapshot

data[]object[]

data[].category"sports" | "politics" | "crypto" | "economy" | "weather" | "entertainment" | "other"

data[].category_source"source" | "inferred" | "unknown"

data[].keystring

data[].venue"polymarket" | "predict_fun"

data[].marketstring

data[].conditionstring

data[].tokenstring

data[].titlestring

data[].outcomestring

data[].urlstring

data[].collateralstring

data[].signal_version"yield.v1" | "yield.v2"

data[].budgetstring

data[].spentstring

data[].unspentstring

data[].sharesstring

Shares received after supported entry share fees (Predict yield.v2); before fees for legacy v1.

data[].purchased_sharesstring

Purchased shares before entry share fees, present in yield.v2.

data[].estimated_entry_fee_sharesstring | null

data[].estimated_entry_feestring | null

Collateral equivalent at entry price. Predict v2 deducts the fee from shares, not as additional cash.

data[].payout_before_entry_fee_if_winstring

data[].net_basisstring

data[].scheduleobject

data[].schedule.sourcestring

data[].schedule.record_idstring

data[].schedule.category_slugstring

data[].schedule.marketstring

data[].schedule.conditionstring

data[].schedule.observed_atstring

data[].schedule.available_atstring

data[].schedule.starts_atstring | null

data[].schedule.ends_atstring | null

data[].schedule.statusstring

data[].recorded_capacitystring

data[].average_pricestring | null

data[].quote_atstring

data[].available_atstring

data[].end_atstring | null

data[].quote_status"observed" | "unavailable"

data[].reasons[]string[]

data[].payout_if_winstring

data[].profit_if_win_grossstring

data[].return_if_win_gross_pctnumber | null

data[].profit_if_win_netstring | null

data[].return_if_win_net_pctnumber | null

data[].annualized_net_pctnumber | null

data[].annualized_net_if_7d_late_pctnumber | null

data[].days_to_scheduled_endnumber | null

data[].feeobject

data[].fee.status"unknown" | "zero" | "observed"

data[].fee.sourcestring

data[].fee.observed_atstring | null

data[].fee.model"polymarket-curve-cash.v1" | "predict-min-shares.v1"

data[].fee.ratestring

data[].fee.rate_bpsinteger

data[].fee.exponent1

data[].contextobject

data[].input_ids[]string[]

metaobject

meta.as_ofstring

meta.signal_version"yield.v1" | "yield.v2"

meta.venue"polymarket" | "predict_fun"

meta.status"preparing" | "live" | "stale"

meta.snapshot_idstring | null

meta.published_atstring | null

meta.input_cutoffstring | null

meta.can_filterboolean

meta.can_replayboolean

meta.coverageobject

meta.recordsnumber

meta.has_moreboolean

meta.next_cursorstring | null

Errors

Recent large Polymarket executions

Latest 100 observed wallet trade rows worth at least 10,000 USDC in the last 24 hours. Deduplicated by source trade ID, not economic transaction. A transaction can include multiple fills and counterparties. Outcome prices are last observed trades within 24 hours with timestamps, not live quotes. Unobserved outcome prices remain null. Categories may be inferred. This is not the calibrated Whale Alert v2 signal or evidence of profitability.

Response · 200

Bounded observed trade sample

data[]object[]

data[].idstring

data[].walletstring

data[].marketstring

data[].tokenstring

data[].outcomestring

data[].sidestring

data[].usdnumber

data[].pricenumber

data[].sharesnumber

data[].txstring

data[].timestring

data[].titlestring

data[].category"sports" | "politics" | "crypto" | "economy" | "weather" | "entertainment" | "other"

data[].category_source"source" | "inferred" | "unknown"

data[].options[]object[]

data[].options[].outcomestring

data[].options[].pricenumber | null

data[].options[].price_atstring | null

metaobject

meta.as_ofstring

meta.latest_trade_atstring | null

meta.threshold10000

meta.limit100

meta.venue"polymarket"

Errors

Polymarket source leaderboard

Top 50 entries from the Polymarket Data API leaderboard for a category, period and sort order. PnL and volume retain the source definitions; PnL is not asserted to be realized-only and volume is not invested capital. This is distinct from KnownAt Top Traders v2. retrieved_at records retrieval, not the upstream calculation time.

Parameters

categorystring · default OVERALLoptional

Source leaderboard category; OVERALL includes all categories.

periodstring · default WEEKoptional

Source leaderboard period: DAY, WEEK, MONTH or ALL.

sortstring · default PNLoptional

Rank by source PnL or traded volume; neither is a KnownAt skill score.

Response · 200

Validated source leaderboard

data[]object[]

data[].rankstring

data[].proxyWalletstring

data[].userNamestring

data[].volnumber

data[].pnlnumber

metaobject

meta.source"Polymarket Data API"

meta.retrieved_atstring

meta.category"OVERALL" | "SPORTS" | "POLITICS" | "CRYPTO" | "CULTURE" | "WEATHER" | "ECONOMICS" | "TECH" | "FINANCE"

meta.period"DAY" | "WEEK" | "MONTH" | "ALL"

meta.sort"PNL" | "VOL"

meta.limit50

Errors

Public preview: Compare conditional carry returns and token-matched buying context

Immutable yield.v2 input snapshots, with original yield.v1 replay, for Polymarket and Predict.fun. Calculates purchases against recorded ask depth for an entry budget. Payout assumes the selected outcome wins and pays one collateral unit per share; price is not a calibrated probability. Partial depth reports unspent capital. Polymarket entry-fee estimates use the stored feeSchedule with exponent 1 and the pinned SDK cash-fee curve; zero-fee flags are supported. Predict v2 uses each stored market feeRateBps and deducts the base taker fee from purchased shares: floor(bps*min(p,1-p)*shares/p/10000), using D18 price and rounding per observed level. Account-specific discounts and refunds are excluded. purchased_shares and payout_before_entry_fee_if_win are before the share fee; shares and payout_if_win are after it. estimated_entry_fee is the entry-price collateral equivalent of the fee, not an extra cash debit for Predict. Exact per-fill rounding can differ for aggregated depth. Unverified fees remain null. The budget includes known entry fees; gas, funding, withdrawal, redemption and optional builder costs are excluded. Predict scheduled ends come from captured category endsAt with exact market/condition membership and a 15-minute freshness limit; missing dates remain unknown. Scheduled end is not promised settlement. Annualization is simple, not compounded; Predict collateral yield/rewards are excluded. Published Whale Alert and Accumulation context matches venue, condition and token at the snapshot cutoff; purchases are not proof of current holdings. Polymarket Top Traders context uses the last published holdings from top 100 portfolios per collateral, at most 24 hours old, and never claims a current balance. Unreleased rankings remain preparing. Predict holdings stay unavailable until their CTF contract identity is supported. Snapshots refresh every five minutes, cover a bounded candidate sample, and retain publication history for replay. An omitted token in a later snapshot means it is no longer in that candidate sample, not that a prior trade was reversed. Scopes are polymarket:read or predict_fun:read by venue.

Parameters

venuestring · default polymarketoptional

Venue and collateral denomination.

categorystring · default alloptional

Filter by source or inferred market category; all disables category filtering.

budgetstring · default 100optional

Entry budget in venue collateral units, including verified entry fees; unverified fees and external costs are excluded. Default 100; custom budget requires sign-in.

min_return_pctnumbernull · default 0optional

Minimum conditional gross return percentage. Not an estimated win probability.

max_daysintegeroptional

Maximum days until scheduled end; unknown/overdue dates are excluded when set.

settlement_delay_daysintegernull · default 0optional

Additional assumed settlement delay for simple annualization.

contextstring · default anyoptional

Require available, token-matched published buying context. Unreleased indicators never count as zero activity.

sortstring · default returnoptional

Sort by conditional gross return, net return (unknown last), recorded capacity or scheduled end.

as_ofstringoptional

Publication cutoff for point-in-time replay; subscription required for historical cutoffs.

limitinteger · default 20optional

Page size; public access is capped at 20.

cursorstringoptional

Cursor binds venue, query, snapshot and cutoff. Use unchanged filters on subsequent pages.

Response · 200

Calculated carry page from a fixed stored input snapshot

data[]object[]

data[].category"sports" | "politics" | "crypto" | "economy" | "weather" | "entertainment" | "other"

data[].category_source"source" | "inferred" | "unknown"

data[].keystring

data[].venue"polymarket" | "predict_fun"

data[].marketstring

data[].conditionstring

data[].tokenstring

data[].titlestring

data[].outcomestring

data[].urlstring

data[].collateralstring

data[].signal_version"yield.v1" | "yield.v2"

data[].budgetstring

data[].spentstring

data[].unspentstring

data[].sharesstring

Shares received after supported entry share fees (Predict yield.v2); before fees for legacy v1.

data[].purchased_sharesstring

Purchased shares before entry share fees, present in yield.v2.

data[].estimated_entry_fee_sharesstring | null

data[].estimated_entry_feestring | null

Collateral equivalent at entry price. Predict v2 deducts the fee from shares, not as additional cash.

data[].payout_before_entry_fee_if_winstring

data[].net_basisstring

data[].scheduleobject

data[].schedule.sourcestring

data[].schedule.record_idstring

data[].schedule.category_slugstring

data[].schedule.marketstring

data[].schedule.conditionstring

data[].schedule.observed_atstring

data[].schedule.available_atstring

data[].schedule.starts_atstring | null

data[].schedule.ends_atstring | null

data[].schedule.statusstring

data[].recorded_capacitystring

data[].average_pricestring | null

data[].quote_atstring

data[].available_atstring

data[].end_atstring | null

data[].quote_status"observed" | "unavailable"

data[].reasons[]string[]

data[].payout_if_winstring

data[].profit_if_win_grossstring

data[].return_if_win_gross_pctnumber | null

data[].profit_if_win_netstring | null

data[].return_if_win_net_pctnumber | null

data[].annualized_net_pctnumber | null

data[].annualized_net_if_7d_late_pctnumber | null

data[].days_to_scheduled_endnumber | null

data[].feeobject

data[].fee.status"unknown" | "zero" | "observed"

data[].fee.sourcestring

data[].fee.observed_atstring | null

data[].fee.model"polymarket-curve-cash.v1" | "predict-min-shares.v1"

data[].fee.ratestring

data[].fee.rate_bpsinteger

data[].fee.exponent1

data[].contextobject

data[].input_ids[]string[]

metaobject

meta.as_ofstring

meta.signal_version"yield.v1" | "yield.v2"

meta.venue"polymarket" | "predict_fun"

meta.status"preparing" | "live" | "stale"

meta.snapshot_idstring | null

meta.published_atstring | null

meta.input_cutoffstring | null

meta.can_filterboolean

meta.can_replayboolean

meta.coverageobject

meta.recordsnumber

meta.has_moreboolean

meta.next_cursorstring | null

Errors

Public preview: Read published Whale Alert v2 revisions

Immutable emission, correction and retraction stream ordered by publication time (newest first). Uses stored whale-alert.v2 decisions, never recalculates historical alerts. as_of freezes publication visibility; next_cursor preserves that snapshot and filters. Apply revisions by alert_key and revision, not transaction time. A minimum-notional filter retains subsequent retractions of matching alerts. Returns preparing until the explicit public release. Requires polymarket:read or predict_fun:read according to venue. Polymarket and Predict.fun via the venue parameter, with independent calibration and release. Predict.fun uses whale-alert.predict-fun.v2 with identical numerical rules and BNB settlement evidence; unusual position building does not establish smart-money identity or profitability.

Parameters

venuestringoptional

Venue. Defaults to polymarket; predict_fun has its own calibration, immutable source manifest and release gate.

as_ofstringoptional

Publication cutoff. Omit for current visibility; pagination freezes this instant.

limitinteger · default 20optional

Maximum published revisions per page, between 1 and 100.

cursorstringoptional

Exact next_cursor string from the previous page, using unchanged filters and cutoff.

walletstringoptional

Exact lowercase wallet address.

marketstringoptional

Exact lowercase condition ID on the selected venue.

min_net_usdnumbernulloptional

Minimum net buying of a published alert. Later corrections and withdrawals of matching alerts remain visible.

Response · 200

Stored revisions or explicit preparing state.

data[]object[]

data[].alert_keystring

data[].revisioninteger

data[].action"emitted" | "corrected" | "retracted"

data[].published_atstring

data[].input_cutoffstring

data[].calibration_fingerprintstring

data[].payloadobject

data[].payload.candidateobject

data[].payload.candidate.venuestring

data[].payload.candidate.walletstring

data[].payload.candidate.marketstring

data[].payload.candidate.tokenstring

data[].payload.candidate.startstring

data[].payload.candidate.endstring

data[].payload.candidate.buyUsdstring

data[].payload.candidate.sellUsdstring

data[].payload.candidate.netUsdstring

data[].payload.candidate.netSharesstring

data[].payload.candidate.availableAtstring

data[].payload.candidate.executionIds[]string[]

data[].payload.decisionstring

data[].payload.reasons[]string[]

data[].payload.smart_money"unverified"

data[].payload.wallet_profilestring

data[].payload.concurrent_outcome_tokens[]string[]

data[].payload.decision_fingerprintstring

metaobject

meta.status"preparing" | "live"

meta.signal_version"whale-alert.v2" | "whale-alert.predict-fun.v2"

meta.as_ofstring

meta.recordsinteger

meta.has_moreboolean

meta.next_cursorstring | null

meta.first_windowstring | null

meta.can_filterboolean

meta.can_replayboolean

Errors

Public preview: Read published Accumulation v2 revisions

Immutable emission, correction and retraction stream ordered by publication time (newest first). Uses stored accumulation.v2 decisions, never recalculates historical alerts. as_of freezes publication visibility; next_cursor preserves that snapshot and filters. Apply revisions by alert_key and revision, not transaction time. A minimum-notional filter retains subsequent retractions of matching alerts. Returns preparing until the explicit public release. Polymarket and Predict.fun use the same numerical rules over 24h, 72h and 168h. Select venue=predict_fun for BNB evidence and accumulation.predict-fun.v2 history. Requires predict_fun:read or polymarket:read according to venue; sustained buying does not establish smart-money identity or profitability.

Parameters

venuestringoptional

Venue. Defaults to polymarket. Predict.fun has independent BNB evidence, publication history and release.

as_ofstringoptional

Publication cutoff. Omit for current visibility; pagination freezes this instant.

limitinteger · default 20optional

Maximum published revisions per page, between 1 and 100.

cursorstringoptional

Exact next_cursor string from the previous page, using unchanged filters and cutoff.

walletstringoptional

Exact lowercase wallet address.

marketstringoptional

Exact lowercase condition ID on the selected venue.

min_net_usdnumbernulloptional

Minimum net buying of a published alert. Later corrections and withdrawals of matching alerts remain visible.

Response · 200

Stored revisions or explicit preparing state.

data[]object[]

data[].alert_keystring

data[].revisioninteger

data[].action"emitted" | "corrected" | "retracted"

data[].published_atstring

data[].input_cutoffstring

data[].calibration_fingerprintstring

data[].payloadobject

data[].payload.candidateobject

data[].payload.candidate.venuestring

data[].payload.candidate.walletstring

data[].payload.candidate.marketstring

data[].payload.candidate.tokenstring

data[].payload.candidate.startstring

data[].payload.candidate.endstring

data[].payload.candidate.asOfstring

data[].payload.candidate.netUsdstring

data[].payload.candidate.netSharesstring

data[].payload.candidate.boughtUsdstring

data[].payload.candidate.soldUsdstring

data[].payload.candidate.marketBoughtUsdstring

data[].payload.candidate.marketBuyShareBpsnumber

data[].payload.candidate.buyCountnumber

data[].payload.candidate.buyHoursnumber

data[].payload.candidate.buySpanMsnumber

data[].payload.candidate.executions[]object[]

data[].payload.candidate.executions[].idstring

data[].payload.candidate.executions[].venuestring

data[].payload.candidate.executions[].walletstring

data[].payload.candidate.executions[].marketstring

data[].payload.candidate.executions[].tokenstring

data[].payload.candidate.executions[].side"BUY" | "SELL"

data[].payload.candidate.executions[].eventAtstring

data[].payload.candidate.executions[].availableAtstring

data[].payload.candidate.executions[].usdstring

data[].payload.candidate.executions[].sharesstring

data[].payload.decisionstring

data[].payload.reasons[]string[]

data[].payload.decision_fingerprintstring

metaobject

meta.status"preparing" | "live"

meta.signal_version"accumulation.v2" | "accumulation.predict-fun.v2"

meta.as_ofstring

meta.recordsinteger

meta.has_moreboolean

meta.next_cursorstring | null

meta.first_windowstring | null

meta.can_filterboolean

meta.can_replayboolean

Errors

Public preview: Read a verified Top Traders ranking snapshot

Polymarket and Predict.fun have independent evidence, versioned snapshots and public release gates. Select venue=predict_fun with predict_fun:read; defaults to polymarket with polymarket:read. Predict.fun uses top-traders.predict-fun.v2 and stays preparing until wallet costs, fee treatment and settlement are reconciled. Seven-day realized execution P&L after charged execution fees. Rankings are separate for each collateral token; amounts are nominal collateral units, not an FX-adjusted USD total. Rewards, fee rebates and unrealized price changes are excluded. ROI is return on realized cost, not capital at risk. At least five traded markets must be fully closed and realized cost must be at least 100 collateral units. Unknown cost or incomplete portfolio evidence excludes the portfolio, with coverage counts reported. as_of and next_cursor freeze the published snapshot and its revision. Corrections never rewrite earlier snapshots. Returns preparing until a validated public release. No sampled-position audit is eligible for this feed.

Parameters

venuestringoptional

Venue; defaults to polymarket. Predict.fun remains preparing until complete wallet cost and settlement evidence is verified.

as_ofstringoptional

Publication cutoff; omit for latest visibility. Historical replay requires an eligible subscription.

limitinteger · default 20optional

Maximum portfolios per page. Anonymous web access is limited to 20.

cursorstringoptional

Exact next_cursor from the previous page; preserves snapshot, revision, filters and publication cutoff.

walletstringoptional

Exact lowercase wallet address. Sign-in required.

collateralstringoptional

Exact lowercase collateral contract. Rankings and nominal amounts are separate per collateral.

Response · 200

Immutable ranking snapshot or preparing state

data[]object[]

data[].walletstring

data[].collateralstring

data[].rankinteger

data[].realized_pnlstring

data[].realized_coststring

data[].closed_marketsinteger

data[].return_on_realized_cost_bpsnumber

data[].open_positions[]object[]

data[].open_positions[].tokenstring

data[].open_positions[].sharesstring

data[].open_positions[].coststring | null

data[].proof_hashstring

metaobject

meta.status"preparing" | "live"

meta.signal_version"top-traders.v2" | "top-traders.predict-fun.v2"

meta.as_ofstring

meta.recordsinteger

meta.next_cursorstring | null

meta.has_moreboolean

meta.can_filterboolean

meta.can_replayboolean

meta.snapshot_keystring | null

meta.revisioninteger | null

meta.window_startstring | null

meta.window_endstring | null

meta.observed_portfoliosinteger

meta.eligible_portfoliosinteger

meta.excluded_portfoliosinteger

Errors

BTC liquidation map research alpha

Returns the fixed Bybit BTCUSDT research snapshot used in the dashboard, including scenario densities and original input hash. This artifact is not live, does not update on request, and does not identify actual liquidation positions. Read eventEnd and generatedAt before interpreting it. Free authenticated access with datasets:read.

Parameters

symbolstring · default BTCoptional

Research alpha currently covers BTC only.

Response · 200

Dated research artifact

dataobject

data.versionstring

data.generatedAtstring

data.eventEndstring

data.marketstring

data.inputSha256string

data.historyHoursnumber

data.missingHoursnumber

data.scenarios[]object[]

data.scenarios[].idstring

data.scenarios[].labelstring

data.scenarios[].coverageobject

data.scenarios[].coverage.longnumber

data.scenarios[].coverage.shortnumber

data.scenarios[].rows[]object[]

data.scenarios[].rows[].timestring

data.scenarios[].rows[].pricenumber

data.scenarios[].rows[].cells[]number[][]

metaobject

meta.status"research_snapshot"

meta.livefalse

Errors

Macro indicators with their latest values

Inflation, labor, growth, consumer activity, rates and energy series with the latest value, a headline (level, year-over-year percent change or change on the previous period, as stated by basis), up to 24 recent headline points and the next scheduled release. Values come from the official publishers (BLS, BEA, DOL, New York Fed, Treasury, EIA; ONS for Revolut card spending, Eurostat, TSA) as they were captured; every change is kept with the time it became known, so revisions stay visible. capture_started gives when each source was first captured: for earlier periods the first value is our first capture, not the original release. sources reports when each publisher was last checked.

Response · 200

Every curated macro series.

data[]object[]

data[].idstring

data[].namestring

data[].group"inflation" | "labor" | "growth" | "consumer" | "rates" | "energy"

data[].frequency"daily" | "weekly" | "monthly" | "quarterly"

data[].unitstring

data[].basis"level" | "yoy" | "change"

level: the value; yoy: percent change on a year earlier; change: difference from the previous period.

data[].sourcestring

data[].source_labelstring

data[].source_urlstring

data[].releasestring | null

data[].notestring | null

What the publisher says the numbers do not show, when it says so.

data[].capture_startedstring | null

When this source was first captured; earlier periods show first capture, not first release.

data[].latestobject | null

data[].latest.periodstring

data[].latest.datestring | null

data[].latest.valuenumber | null

data[].latest.first_seen_atstring

data[].latest.last_changed_atstring

data[].latest.versionsnumber

data[].previousobject | null

data[].previous.periodstring

data[].previous.valuenumber | null

data[].headlineobject | null

data[].headline.basis"level" | "yoy" | "change"

data[].headline.valuenumber

data[].headline.periodstring

data[].headline.compare_periodstring | null

data[].spark[]object[]

data[].spark[].periodstring

data[].spark[].valuenumber | null

data[].next_releaseobject | null

data[].next_release.namestring

data[].next_release.scheduled_atstring

sources[]object[]

sources[].sourcestring

sources[].last_checkstring

sources[].last_successstring

sources[].outcomestring

metaobject

meta.generated_atstring

meta.capture_startedstring

meta.notestring

meta.groups[]string[]

Errors

History and revisions of one macro series

One series, newest periods first up to limit, each with its current value, the first value captured and when, and the number of versions. revisions lists every change of a published value, newest first. as_of returns what was known at that instant. Values come from the official publishers (BLS, BEA, DOL, New York Fed, Treasury, EIA; ONS for Revolut card spending, Eurostat, TSA) as they were captured; every change is kept with the time it became known, so revisions stay visible. capture_started gives when each source was first captured: for earlier periods the first value is our first capture, not the original release.

Parameters

seriesstringrequired

Series ID from /v1/macro/series, for example cpi, payrolls, gdp, sofr or crude_stocks.

as_ofstringoptional

Knowledge cutoff: only versions captured at or before this instant, so a backtest sees what was known then. Defaults to now.

limitinteger · default 120optional

Most recent periods to return.

Response · 200

Periods and revisions of the series.

seriesobject

series.idstring

series.namestring

series.group"inflation" | "labor" | "growth" | "consumer" | "rates" | "energy"

series.frequency"daily" | "weekly" | "monthly" | "quarterly"

series.unitstring

series.basis"level" | "yoy" | "change"

level: the value; yoy: percent change on a year earlier; change: difference from the previous period.

series.sourcestring

series.source_labelstring

series.source_urlstring

series.releasestring | null

series.notestring | null

What the publisher says the numbers do not show, when it says so.

observations[]object[]

observations[].periodstring

observations[].datestring | null

observations[].valuenumber | null

observations[].first_valuenumber | null

observations[].first_seen_atstring

observations[].last_changed_atstring

observations[].versionsnumber

observations[].first_releasestring | null

observations[].headlinenumber | null

revisions[]object[]

revisions[].periodstring

revisions[].fromnumber | null

revisions[].tonumber | null

revisions[].atstring

revisions[].releasestring | null

metaobject

meta.as_ofstring

meta.periodsnumber

meta.truncatedboolean

meta.capture_startedstring

meta.notestring

Errors

Scheduled macro releases and capture latency

Releases from the BLS, BEA and ONS calendars, EIA and DOL weekly reports and the daily New York Fed and Treasury publications. Around each one the collector polls every 2 seconds; captured releases report the delay from the scheduled time to the first change seen, and the series each release updates.

Parameters

fromstringoptional

Inclusive start. Defaults to 7 days ago.

tostringoptional

Inclusive end, at most 120 days after from. Defaults to 14 days ahead.

Response · 200

Releases in the window, oldest first.

data[]object[]

data[].sourcestring

data[].keystring

data[].namestring

data[].scheduled_atstring

data[].probes[]string[]

data[].series[]string[]

data[].status"upcoming" | "in_window" | "captured" | "no_capture"

data[].first_change_atstring | null

data[].latency_msnumber | null

metaobject

meta.fromstring

meta.tostring

meta.generated_atstring

meta.window_minutesnumber

meta.notestring

Errors