generated: '2026-07-18'
method: searched
source: https://inference-docs.cerebras.ai/support/rate-limits
units: [RPM, RPH, RPD, TPM, TPH, TPD]
model: >-
  Whichever metric is exceeded first triggers limiting. Organizations have two
  independent token buckets: uncached TPM (full-compute tokens) and total TPM
  (typically 3x the uncached limit, includes cached tokens). Exceeding returns
  429 Too Many Requests indicating which bucket was hit.
exceeded_status: 429
tiers:
- tier: Free Trial
  limits:
  - model: gpt-oss-120b
    limit_count: 5
    rpm: 5
    tpm: 30000
    tph: 1000000
    tpd: 1000000
  - model: zai-glm-4.7
    rpm: 5
    tpm: 30000
    tph: 1000000
    tpd: 1000000
  - model: gemma-4-31b
    rpm: 5
    tpm: 30000
    tph: 1000000
    tpd: 1000000
- tier: Developer (Pay as You Go)
  note: No hourly/daily restrictions.
  limits:
  - model: gpt-oss-120b
    rpm: 1000
    tpm: 1000000
  - model: zai-glm-4.7
    rpm: 500
    tpm: 500000
  - model: gemma-4-31b
    rpm: 300
    tpm: 500000
- tier: Enterprise
  note: Custom limits based on organization profile.