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: Lamini
providerId: lamini
created: '2026-06-20'
modified: '2026-06-20'
reconciled: false
tags:
- AI
- LLM
- Fine-Tuning
- Memory Tuning
- Inference
- Rate Limiting
- Quotas
- Throttling
description: >-
  The Lamini Platform meters inference and tuning usage per account and is
  governed primarily by available credit / spend on the On-Demand tier and by
  reserved GPU capacity on Enterprise. Concurrent inference requests and
  tuning jobs are bounded by account capacity rather than fixed published
  per-minute request quotas. Specific numeric limits are not publicly
  documented and are not reconciled in this artifact.
notes: >-
  Lamini does not publish a fixed per-minute request/token rate table the way
  OpenAI-style providers do; throughput is bounded by On-Demand credit and
  Enterprise reserved capacity. Confirm any concurrency or job limits with
  Lamini support during reconciliation.
sources:
- https://docs.lamini.ai
- https://www.lamini.ai/pricing
responseCodes:
  throttled: 429
limits:
- name: Concurrent Inference Requests
  scope: account
  metric: requests
  limit: see provider documentation
  notes: Inference concurrency bounded by account capacity and credit/tier.
- name: Concurrent Tuning Jobs
  scope: account
  metric: jobs
  limit: see provider documentation
  notes: Number of simultaneous tuning jobs bounded by GPU capacity / tier.
- name: Tuning Throughput
  scope: account
  metric: steps
  limit: see provider documentation
  notes: Burst tuning scales linearly across multiple GPUs/nodes on On-Demand.
- name: Credit / Spend Ceiling
  scope: account
  metric: usd
  limit: see provider documentation
  notes: On-Demand usage is bounded by available prepaid or free credit.
policies:
- name: Capacity-Based Limits
  description: Throughput scales with On-Demand credit and Enterprise reserved GPU capacity rather than fixed per-minute quotas.
- name: Backoff Strategy
  description: Clients should implement exponential backoff with jitter and honor Retry-After on 429 responses.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com