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: Sakana AI
providerId: sakana-ai
created: '2026-05-08'
# Provenance stamped 2026-08-11: this artifact was written by the API Evangelist
# bulk sweep dated 2026-05-08, not harvested from the provider. See roadmap#35.
method: generated
modified: '2026-05-08'
reconciled: false
tags:
- AI
- LLM
- Research
- Multi-Agent
- Rate Limiting
description: >-
  Sakana Fugu rate limits are not published; service is in beta. Apply general OpenAI-
  compatible client patterns (exponential backoff, Retry-After).
notes: Update once Fugu publishes a public rate-limit policy.
sources:
- https://sakana.ai/
responseCodes:
  throttled: 429
limits:
- name: Beta Default
  scope: account
  metric: requests-per-minute
  limit: see beta enrollment
  notes: Allocated per beta program participant.
policies:
- name: Backoff Strategy
  description: Exponential backoff with jitter; honor Retry-After.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com