Pricing Catalog API

Pricing Catalog API

GET /x-pricing

Returns the pricing catalog. The price table itself is public data and can be read anonymously; with an API Key, the final prices for that account are returned.

Request

ModeTenant of the catalogMultiplier
With Authorization: Bearer <key>The account's tenantThe account's rates × factor
Anonymous, with ?oid=<tenant ID> (or ?owner=, ?public_key=pk-Xvs...)The specified tenant1
Anonymous, no parametersThe default tenant1
# Public catalog
curl "$BASE_URL/x-pricing"

# Final prices for the current account
curl "$BASE_URL/x-pricing" -H "Authorization: Bearer $API_KEY"
  • public_key is the account's public identifier, available from GET /dashboard/status; a malformed value returns 400 invalid public_key, and an unknown one returns 404 public key not found. oid and owner must be positive integers, otherwise 400 owner id must be a positive integer.
  • When Authorization is sent but the key is invalid, 401 is returned directly; it does not fall back to the public catalog.
  • Only GET is accepted; OPTIONS returns 204; other methods return 405 method not allowed.
  • The response carries Cache-Control: no-cache.

Response

{
  "object": "pricing_catalog",
  "oid": 1,
  "public": false,
  "currency": "USD",
  "unit_notes": {
    "per_1m_tokens": "Token prices are displayed per 1M tokens.",
    "per_minute": "Voice duration is displayed per minute and billed per second.",
    "per_image": "Image size prices are displayed per generated image.",
    "per_second": "Video resolution prices are displayed per generated second.",
    "per_call": "Call prices are displayed per request."
  },
  "generated_at": "2026-11-20T00:00:00Z",
  "user": {"id": 42, "oid": 1, "rates": 1, "factor": 1, "multiplier": 1},
  "data": [
    {
      "id": "gpt-6.2-sol",
      "category": "ChatPricing",
      "kind": "chat",
      "unit": "per_1m_tokens",
      "base": {"InputText": 2, "CachedText": 0.1, "CacheWrite": 2.5, "OutputText": 10, "FastTokenMultiplier": 2, "Rates": 1},
      "effective": {"InputText": 2, "CachedText": 0.1, "CacheWrite": 2.5, "OutputText": 10, "FastTokenMultiplier": 2, "Rates": 1},
      "final": {"InputText": 2, "CachedText": 0.1, "CacheWrite": 2.5, "OutputText": 10, "FastTokenMultiplier": 2},
      "rates": {"model": 1, "user": 1, "factor": 1, "effective": 1},
      "override": false,
      "custom": false
    }
  ]
}

The example omits the long-context tiers ContextTiers; the price fields each model actually carries are those in the live response.

FieldDescription
userPresent only when a key is sent
data[].category, kindPrice table category: ChatPricing/chat, ImgPricing/image, VideoPricing/video, AudioPricing/audio, RerankPricing/rerank, CallPricing/call, FineTuningPricing/fine_tuning
data[].unitper_1m_tokens, per_image, per_second, per_minute, per_call
baseSystem default price
effectivePrice after tenant custom pricing is applied
finalFinal price: price field × model Rates (long-context tiers may carry their own Rates) × account rates × factor; the Call of chat, fine-tuning, and rerank is multiplied by the account multiplier only; InputThreshold and FastTokenMultiplier keep their original values; no Rates field
ratesModel multiplier, account multiplier, price factor, and their product
overrideWhether it differs from the system default
customWhether the model was added by the tenant

currency is the catalog's base currency USD; accounts whose price factor is not 1 are converted by that factor and settle in yuan. The token prices in final do not yet include the speed-tier (fast/ultrafast) multiplier; see Router API.