Source: Vertex AI — Consumption Options
The five ways to pay for Gemini on Vertex AI: Standard, Priority and Flex PayGo, Provisioned Throughput and Batch.
Contents
- What: vendor documentation
- Author: Google Cloud
- Date: undated, read 2026-11-14
- Original: Consumption options
Summary
Compares the consumption options on price, reliability and throttling, and suggests how to combine them.
Key claims
- Standard PayGo: standard rate, no QoS guarantee, subject to 429s.
- Priority PayGo: premium rate, priority processing, with a ramp limit.
- Flex PayGo: about 50% cheaper, slower, more throttling; for latency-tolerant work.
- Provisioned Throughput: terms of 1 week, 1 month, 3 months or 1 year; a 1-year term is about 26% cheaper than 1 month.
- Batch: about 50% cheaper than Standard PayGo, asynchronous.
- Suggested pattern: PT for steady baseline, spill to Standard/Priority PayGo, Batch for backlogs, Flex for tolerant tasks; use exponential backoff and the global endpoint against 429s.