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: Zhipu AI
providerId: zhipu-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: true
tags:
- AI
- LLM
- GLM
- Rate Limiting
- Quotas
description: >-
  Z.ai applies model-specific concurrency to API users with prepaid balance. GLM Coding Plan
  subscribers have 5-hour and weekly usage windows with peak/off-peak quota multipliers.
notes: >-
  Z.ai does not publish a comprehensive numeric RPM/TPM/concurrency table; concurrency is
  managed per-model and per-account. Some paid tiers default to 1 in-flight request unless
  the package explicitly grants more.
sources:
- https://z.ai/manage-apikey/rate-limits
- https://docs.z.ai
responseCodes:
  throttled: 429
limits:
- name: API Default Concurrency
  scope: account
  metric: concurrent-requests
  limit: 1
  notes: Default per-model concurrency for prepaid API users; varies by package.
- name: GLM Coding Plan Window
  scope: account
  metric: window
  limit: '5-hour and weekly'
  notes: Usage allowances reset on rolling 5-hour and weekly windows.
- name: Peak Hours Multiplier
  scope: account
  metric: quota-multiplier
  limit: '2-3x'
  notes: Peak (14:00-18:00 UTC+8) consumes quota at 2-3x the off-peak rate for GLM-5.1 / GLM-5-Turbo.
policies:
- name: Backoff Strategy
  description: Exponential backoff with jitter; honor Retry-After headers.
- name: Concurrency Upgrade
  description: Higher concurrency requires contacting Z.ai sales or upgrading plan tier.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com