Open Dev Kit Documentation :: Open AI :: AI Model Service Tier

OpenAI.AIModelServiceTier

Constants

Auto (`auto`)

Uses the OpenAI project’s configured service tier. On Scale Tier–enabled projects this consumes Scale Tier capacity while available; otherwise it behaves like 'default'.

Default (`default`)

Uses the standard processing tier (normal pricing and availability, with no special latency or uptime guarantees).

Flex (`flex`)

Uses the Flex processing tier: lower per-token cost with higher and more variable latency; requests may queue or be rate-limited during busy periods.

Priority (`priority`)

Uses the Priority processing tier: higher per-token cost for faster and more consistent latency, especially under load.

If you think anything is missing, please feel free to: submit documentation feedback on this page