generated: '2026-08-17'
method: searched
source: https://sifting.io/docs/errors
docs: https://sifting.io/docs/errors
also: [https://sifting.io/pricing.md, 'openapi/_original/siftingio-openapi.yaml (components.responses.TooManyRequests)']
note: >-
  SiftingIO publishes a dedicated "Errors & rate limits" operations page that gives an agent
  everything it needs at runtime: the exact response headers, the exhaustion status code, the
  error code to switch on, and the numeric budgets per tier. Two INDEPENDENT limits apply — a
  monthly call quota and a per-second (Free: per-minute) rate limit — and both are enforced PER
  MARKET at that market's tier, so one account can run Stocks on Pro and Crypto on Builder with
  separately enforced budgets. The 429 + Retry-After contract is also present in the OpenAPI:
  components.responses.TooManyRequests declares the Retry-After header, and every one of the 37
  operations declares a 429 response.

headers:
  - name: X-RateLimit-Limit
    meaning: Token bucket burst capacity for your subscription tier.
    on: every REST response
  - name: X-RateLimit-Remaining
    meaning: Tokens remaining in the current window.
    on: every REST response
  - name: Retry-After
    meaning: Seconds to wait before retrying.
    on: 429 responses
    in_openapi: true
    openapi_ref: components.responses.TooManyRequests.headers.Retry-After (integer)
header_style: X-RateLimit-* (de-facto draft form, not the RFC 9331 RateLimit-* field names)
self_throttle_guidance: >-
  "Every REST response carries these. Read them to self-throttle before you hit a 429." —
  https://sifting.io/docs/errors

exhaustion:
  status: 429
  error_code: rate_limit_exceeded
  error_meaning: Per-tier rate budget exhausted, inspect Retry-After.
  body_field: retry_after
  body_field_source: components.schemas.Error.retry_after ("Present on rate-limit errors.")
  envelope: '{"error":"rate_limit_exceeded","message":"...","retry_after":<seconds>}'

limits:
  - scope: per-api-key, per-market
    tier: Free
    window: month
    limit: 10000
    unit: REST calls
  - scope: per-api-key, per-market
    tier: Free
    window: minute
    limit: 60
    unit: requests
  - scope: per-api-key, per-market
    tier: Builder
    window: month
    limit: 250000
    unit: REST calls
  - scope: per-api-key, per-market
    tier: Builder
    window: second
    limit: 100
    unit: requests
  - scope: per-api-key, per-market
    tier: Pro
    window: month
    limit: 5000000
    unit: REST calls
  - scope: per-api-key, per-market
    tier: Pro
    window: second
    limit: 150
    unit: requests
  - scope: per-api-key, per-market
    tier: Ultra
    window: month
    limit: null
    unit: REST calls
    note: Unlimited calls.
  - scope: per-api-key, per-market
    tier: Ultra
    window: second
    limit: 250
    unit: requests
    burst: upgradeable in +200 req/sec increments as a paid add-on
  - scope: per-account
    tier: Enterprise
    window: negotiated
    limit: null
    unit: custom
    note: Custom quota and rate, cross-market.

websocket_limits:
  - {tier: Free, connections: 1, subscriptions: 5}
  - {tier: Builder, connections: 3, subscriptions: 100}
  - {tier: Pro, connections: 10, subscriptions: 1000}
  - {tier: Ultra, connections: 50, subscriptions: null}
  - {tier: Enterprise, connections: custom, subscriptions: custom}
websocket_note: >-
  The AsyncAPI document (asyncapi/siftingio-asyncapi.yaml) documents no rate-limit frame; the
  stream's control surface is subscribe/unsubscribe/ping with ack/pong/error replies. Quota
  enforcement on the stream is by connection and subscription count, not by message rate.

quota_exhaustion_behaviour:
  free: 'Hard cap. Calls past the monthly quota return 429 with Retry-After until the next billing cycle.'
  builder: 'Hard cap, same as Free — "a hard cap and a predictable bill".'
  pro: 'Opt-in soft overage at a transparent per-call rate.'
  enterprise: 'Negotiated overage and committed throughput.'
  provider_commitment: '"We never silently bill or degrade." — https://sifting.io/docs/errors'

raising_limits:
  - {mechanism: Extra REST calls add-on, detail: 1M-call packs, price_from_usd_month: 79}
  - {mechanism: Rate-limit upgrade add-on, detail: +200 req/sec increments (Pro and Ultra), price_from_usd_month: 199}
  - {mechanism: Extra WebSocket capacity add-on, detail: +5 connections / +500 subscriptions, price_from_usd_month: 59}
stability_commitment: 'Upgrades raise quota and rate limit "without changing the headers or codes".'

limit_count: 9