ReferenceLive V1

API reference

The OpenAI-compatible surface plus the Omnious extensions: analytics, market data, integrity, points, orders.

The customer-facing API is OpenAI-compatible: POST /v1/chat/completions works as a drop-in base_url swap for existing apps and agent frameworks, with images and embeddings alongside. Payment rides the x402 flow. An unpaid request answers 402 Payment Required with an acceptsarray quoting the worst-case cost at the winner's cleared prices; the client signs a USDC authorization and retries the same request with the x402 v2 PAYMENT-SIGNATURE header. Verification takes about 100 ms and inference starts while settlement runs concurrently. Payload details are in Wire formats.

Auth in the table below: pay means an x402 payment signature, pub is a public read, key reads your own data via x-api-key, sig is a wallet typed-data signature, and provider is a quote signed by a registered provider key. Admin surfaces (/v1/admin/*) exist but are operator-token gated and not listed here.

EndpointAuthPurpose
POST /v1/chat/completionspayOpenAI-compatible completions; the core paid surface
POST /v1/images/generationspaypaid image generation
POST /v1/embeddingspayembeddings, billed input-only at the cleared price
POST /v1/lockspaybuy a forward price lock for a model class
POST /v1/orderspayplace a patient limit order
GET /healthpubrouter status, chain network, auction version
GET /v1/modelspubmodel classes with live depth and prices
GET /v1/market/bookpubanonymized ask ladder per class
GET /v1/market/statspubrolling clearing summary (24h window)
GET /v1/market/snapshotpubfull point-in-time market snapshot
GET /v1/market/candlespubcleared-price OHLC and VWAP candles
GET /v1/market/indexpubcomposite price index and component weights
GET /v1/receipts/:idpubthe signed usage record for a request
GET /v1/analytics/requests/:idpubreconstructed auction: rivals, counterfactuals, fee split
GET /v1/analytics/sessions/:idpubper-turn session tape and cache discount captured
GET /v1/analytics/marketpubper-class cleared prices, HHI, latency
GET /v1/analytics/providers/:idpuba provider's served volume and reliability
GET /v1/analytics/dashboardpubbucketed chart feed
GET /v1/analytics/leaderboardpubproviders ranked by cleared volume
GET /v1/model-integrity/claims/:idpuba provider's bonded model claim
GET /v1/model-integrity/challengers/:walletpubapproved-challenger status for a wallet
GET /v1/model-integrity/challenges/quotepubstake terms for challenging one receipt, before signing
POST /v1/model-integrity/challengessigopen a points-staked challenge (approved challengers)
GET /v1/points/:walletpubS1 score: earned bps, tier, streak, rank, challenge record
GET /v1/points/leaderboardpubthe points board
GET /v1/points/provider/:idpubserve-side points for a provider
GET /v1/orders/:idpuborder state: funding, resting, filled, expired
POST /v1/orders/:id/cancelsigcancel a resting order; escrow refunds at epoch close
GET /v1/locks/:idpublock state: remaining budget, expiry
GET /v1/sessions/:idpubsession state and remaining budget
POST /v1/sessions/:id/closesigclose a session; unspent budget refunds
POST /v1/keyssigmint an API key bound to your wallet
GET /v1/keys/usagekeyper-key request and spend attribution
POST /provider/quotesprovidersigned standing-quote ingest (the maker side)
noteProviders do not have to speak OpenAI upstream. The router translates to Anthropic-dialect backends (/v1/messages) and back before the metering tap, so metering, billing, analytics, and the customer never know which dialect the winning backend spoke. Customers always see the OpenAI surface.

Every endpoint here has a typed method in the SDK and most have a CLI command.