Vertex AI — Gemini model configuration and cost/performance overview
Provider: Google Cloud Vertex AI
All extraction requests route through Vertex AI. Multiple Gemini variants are available for different accuracy and latency profiles.
HIGH ACCURACY
Gemini 2.5 Pro
BALANCED
Gemini 2.5 Flash
NEXT-GEN
Gemini 3.0 Pro
Gemini Model Comparison
Gemini 2.5 Pro
Latency: 2.4s · $0.00125/1K in · $0.005/1K out · 1M ctx
Accuracy
98.7%
Gemini 2.5 Flash
Latency: 0.8s · $0.00015/1K in · $0.0006/1K out · 1M ctx
Accuracy
96.2%
Gemini 3.0 Pro
Latency: 2.1s · Pricing TBD · Preview
Accuracy
99.1%
Gemini 3.0 Flash
Latency: 0.5s · Pricing TBD · Preview
Accuracy
97.4%