Initial Quota / Max Quota

Notes

• OpenAI GPT-5 (and later) general models require explicit approval
• xAI grok-4 models require explicit approval
  • grok-4-fast (grok-4-fast-non-reasoning & grok-4-fast-reasoning) are not subject to approval
• Quota is in terms of Tokens Per Minute (TPM)
  • There is also a requests per minute limit. With AI Portal, customers will typical encounter the token limit before hitting the request limit.
• Quota is per Azure subscription per Azure region per model.
• Initial Quota and Max Quota listed are for standard Azure organizations (”pay as you go”, etc). Enterprise customers or customers with a MCA (Microsoft Customer Agreement) in place have higher initial and max quota limits.
• Customers can request Azure to increase their quota for a given model

References

1. https://learn.microsoft.com/en-us/azure/ai-foundry/foundry-models/quotas-limits?view=foundry&preserve-view=true
2. https://learn.microsoft.com/en-us/azure/ai-foundry/openai/quotas-limits?view=foundry-classic&tabs=REST