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: LangDB
providerId: langdb
created: '2026-06-20'
modified: '2026-06-20'
reconciled: false
tags:
- AI
- LLM
- AI Gateway
- Routing
- Governance
- MCP
- Rate Limiting
- Quotas
- Throttling
description: >-
  LangDB enforces project- and user-scoped limits at the gateway. Beyond
  request/token throughput, it offers cost-based controls: per-user and
  per-group spend limits, project budgets, and cost groups (available in
  Business and Enterprise tiers) that throttle or block requests once a
  configured budget is exceeded. Underlying model providers also enforce
  their own rate limits, which the gateway surfaces. Specific numeric
  per-tier limits are not published and are not reconciled in this artifact.
notes: >-
  Confirm per-tier RPM/TPM throughput and the configuration of cost limits,
  project budgets, and cost groups in the LangDB console and cost-control
  docs during reconciliation.
sources:
- https://docs.langdb.ai/features/cost-control
- https://docs.langdb.ai/features/usage
- https://langdb.ai/cost-control/
responseCodes:
  throttled: 429
limits:
- name: Requests Per Minute (RPM)
  scope: project
  metric: requests
  limit: see provider documentation
  notes: Gateway throughput limit; varies by tier and routed provider.
- name: Tokens Per Minute (TPM)
  scope: project
  metric: tokens
  limit: see provider documentation
  notes: Aggregate token throughput; also bounded by the underlying provider's limits.
- name: Per-User Spend Limit
  scope: user
  metric: cost
  limit: configurable
  notes: User-level budget cap (Business / Enterprise); requests blocked when exceeded.
- name: Per-Group Spend Limit
  scope: group
  metric: cost
  limit: configurable
  notes: Group/team budget cap via cost groups (Business / Enterprise).
- name: Project Budget
  scope: project
  metric: cost
  limit: configurable
  notes: Project-level spend cap; throttles or blocks once the budget is reached.
policies:
- name: Cost-Based Throttling
  description: Requests can be blocked or throttled when a user, group, or project budget is exceeded.
- name: Provider Limit Pass-Through
  description: Underlying model provider rate limits apply and are surfaced through the gateway.
- 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