specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Uniblock
providerId: uniblock
created: '2026-05-04'
# Provenance stamped 2026-08-11: this artifact was written by the API Evangelist
# bulk sweep dated 2026-05-04, not harvested from the provider. See roadmap#35.
method: generated
modified: '2026-05-05'
reconciled: true
tags:
  - Blockchain
  - Web3
  - Rate Limiting
description: Uniblock enforces rate limits at the plan level using compute units per second (CU/s)
  along with a monthly CU envelope. Limits are bound to the project's API key. Exceeding the rate
  returns HTTP 429; clients should back off, batch, cache, or upgrade to a higher plan. Specific
  retry headers are not documented publicly.
sources:
  - https://uniblock.dev/pricing
  - https://docs.uniblock.dev/guides/uniblock/rate-limits
responseCodes:
  throttled: 429
  quotaExceeded: 429
limits:
  - name: Startup plan rate
    scope: project
    metric: compute_units_per_second
    limit: 1000
    timeFrame: second
    notes: Free Startup plan; 40M CUs per month.
  - name: Growth plan rate
    scope: project
    metric: compute_units_per_second
    limit: 2000
    timeFrame: second
    notes: Growth plan; 500M CUs per month.
  - name: Pro plan rate
    scope: project
    metric: compute_units_per_second
    limit: 8000
    timeFrame: second
    notes: Pro plan; 2B CUs per month.
  - name: Business plan rate
    scope: project
    metric: compute_units_per_second
    limit: 20000
    timeFrame: second
    notes: Business plan; 5.5B CUs per month.
  - name: Enterprise plan rate
    scope: contract
    metric: compute_units_per_second
    limit: custom
    notes: Negotiated CU/s and dedicated nodes per contract.
policies:
  - name: Backoff Strategy
    description: On HTTP 429, implement retry with exponential backoff or brief delay before re-submission.
  - name: Caching and Batching
    description: Reduce CU consumption by lowering polling frequency, batching requests, and caching
      responses where possible.
  - name: Plan Upgrade
    description: Sustained throughput needs above plan limits should be addressed by upgrading plan
      tier rather than retrying past the limit.
  - name: Plan-Scoped Enforcement
    description: Rate limits are enforced at the Uniblock plan level per project, not globally per
      account.
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com